B-161483, JULY 14, 1967, 47 COMP. GEN. 29

B-161483: Jul 14, 1967

Additional Materials:

Contact:

Edda Emmanuelli Perez
(202) 512-2853
EmmanuelliPerezE@gao.gov

 

Office of Public Affairs
(202) 512-4800
youngc1@gao.gov

WHICH HAVING MET ALL REQUIREMENTS EXCEPT ONE PORTION OF A DEMONSTRATION TEST IS WITHIN A COMPETITIVE RANGE. 1967: FURTHER REFERENCE IS MADE TO THE PROTEST OF HONEYWELL INCORPORATED. WHICH WAS THE SUBJECT OF LETTERS DATED JUNE 7 AND 16. PHASE II IS IMPLEMENTATION OF THE AIR FORCE'S PLAN TO STANDARDIZE DATA SYSTEMS AND DATA PROCESSING EQUIPMENT AT THE BASE LEVEL WHICH WAS INITIATED AS A TWO-PHASE PROGRAM IN 1962. IS REPORTED WELL UNDERWAY. APPARENTLY A THIRD PHASE WAS DECIDED ON SUBSEQUENT TO 1962. PHASE II IS INTENDED TO PROVIDE SUPPORT FOR MANAGEMENT FUNCTIONS NOT INCLUDED IN EITHER OF THE OTHER TWO PHASES. THE DATA PROCESSING INDUSTRY WAS INFORMED OF PLANS TO PROCEED WITH PHASE II AS EARLY AS JANUARY 21.

B-161483, JULY 14, 1967, 47 COMP. GEN. 29

EQUIPMENT - AUTOMATIC DATA PROCESSING SYSTEMS - SELECTION AND PURCHASE - NEGOTIATION PROCEDURES THE REFUSAL OF THE AIR FORCE IN SELECTING A SOURCE FOR FURNISHING ELECTRIC DATA PROCESSING EQUIPMENT (EDPE), TO BE PURCHASED BY GENERAL SERVICES ADMINISTRATION (GSA) UNDER THE FEDERAL SUPPLY SYSTEM, TO DISCUSS THE TECHNICAL DEFICIENCIES OF THE PROPOSAL THAT OFFERED A LOWER PRICE THAN THAT OF THE ONLY PROPOSAL OUT OF FOUR CONSIDERED ACCEPTABLE VIOLATED 10 U.S.C. 2304 (G), WHICH PROVIDES FOR WRITTEN OR ORAL DISCUSSIONS WITH ALL RESPONSIBLE OFFERORS SUBMITTING PROPOSALS WITHIN A COMPETITIVE RANGE, PRICE AND OTHER FACTORS CONSIDERED, WHEN A NEGOTIATED PROCUREMENT EXCEEDS $2,500, AND THE AUTHORITY OF GSA TO COORDINATE AND PROVIDE FOR THE ECONOMIC AND EFFICIENT ACQUISITION OF EDPE NEITHER IMPAIRING THE SELECTION RIGHT OF AN AGENCY NOR EXEMPTING THE SELECTION FROM PROCUREMENT LAWS AND REGULATIONS, FURTHER DISCUSSIONS SHOULD BE CONDUCTED ON THE LOW PROPOSAL, WHICH HAVING MET ALL REQUIREMENTS EXCEPT ONE PORTION OF A DEMONSTRATION TEST IS WITHIN A COMPETITIVE RANGE, AND ON ANY OTHER PROPOSALS SATISFYING THE "WITHIN A COMPETITIVE RANGE" REQUIREMENT.

TO THE SECRETARY OF THE AIR FORCE, JULY 14, 1967:

FURTHER REFERENCE IS MADE TO THE PROTEST OF HONEYWELL INCORPORATED, AGAINST THE SELECTION OF IBM AS THE SOURCE FOR FURNISHING THE STANDARD ELECTRONIC DATA PROCESSING EQUIPMENT TO BE USED IN SUPPORT OF THE AIR FORCE'S BASE LEVEL DATA AUTOMATION STANDARDIZATION PLAN KNOWN AS PHASE II (ESQ PROJECT 25-65), WHICH WAS THE SUBJECT OF LETTERS DATED JUNE 7 AND 16, 1967, WITH SUPPORTING REPORTS AND DOCUMENTS, FROM WILLIAM MUNVES, DEPUTY GENERAL COUNSEL, AND LEONARD MARKS, JR., ASSISTANT SECRETARY, RESPECTIVELY.

PHASE II IS IMPLEMENTATION OF THE AIR FORCE'S PLAN TO STANDARDIZE DATA SYSTEMS AND DATA PROCESSING EQUIPMENT AT THE BASE LEVEL WHICH WAS INITIATED AS A TWO-PHASE PROGRAM IN 1962. PHASE I, ENCOMPASSING CERTAIN MANAGEMENT FUNCTIONS, IS REPORTED WELL UNDERWAY. APPARENTLY A THIRD PHASE WAS DECIDED ON SUBSEQUENT TO 1962, AND THIS HAS BEEN IMPLEMENTED AS PHASE III. PHASE II IS INTENDED TO PROVIDE SUPPORT FOR MANAGEMENT FUNCTIONS NOT INCLUDED IN EITHER OF THE OTHER TWO PHASES, SUCH AS MILITARY AND CIVILIAN PERSONNEL, MAINTENANCE ENGINEERING, CIVIL ENGINEERING, ACCOUNTING AND FINANCE, MEDICAL EQUIPMENT AND SERVICES, OPERATIONS AND SUPPORT SAFETY, AND TRANSPORTATION.

THE DATA PROCESSING INDUSTRY WAS INFORMED OF PLANS TO PROCEED WITH PHASE II AS EARLY AS JANUARY 21, 1966, IN A LETTER FROM THE ELECTRONIC SYSTEMS DIVISION, HANSCOM FIELD, MASSACHUSETTS. THIS WAS FOLLOWED BY A LETTER DATED FEBRUARY 8, 1966, FROM ASSISTANT SECRETARY MARKS TO 32 COMPANIES ADVISING OF HIS INTEREST IN LEARNING INDUSTRY INTENT AND CAPABILITY TO MEET THE NEEDS FOR "THIRD GENERATION" MODULARITY IN BOTH EQUIPMENT AND SOFTWARE. A FURTHER LETTER REQUESTING EXPRESSIONS OF INTEREST IN RECEIVING THE SOON TO BE ISSUED REQUEST FOR PROPOSALS WAS SENT TO 44 COMPANIES ON MAY 29, 1966. REQUEST FOR PROPOSALS (RFP) NO. ESQ PROJECT 25 -65 WAS ISSUED ON JULY 31, 1966, TO THE 15 COMPANIES EXPRESSING AN INTEREST IN RECEIVING IT. THE RFP, AS AMENDED, CALLED FOR SUBMISSION OF PROPOSALS IN THREE SEPARATELY BOUND BINDERS, WITH PART I (SYSTEMS PROPOSAL) AND PART II (TECHNICAL DATA) DUE ON DECEMBER 5, 1966, AND PART III (COST DATA) DUE ON DECEMBER 30, 1966. ORAL PRESENTATION OF PROPOSALS WAS SCHEDULED TO BEGIN ON DECEMBER 6, 1966. LIVE TEST DEMONSTRATIONS OF BENCHMARK PROBLEMS USING THE PROPOSED EQUIPMENT CONFIGURATIONS AND SOFTWARE WERE SCHEDULED TO BEGIN ON JANUARY 16, 1967. ANNOUNCEMENT OF THE SELECTION OF THE SUCCESSFUL OFFEROR WAS SCHEDULED TO BE MADE WITHIN 120 DAYS FROM THE CLOSING DATE FOR SUBMISSION OF PROPOSALS. THE FOREGOING SCHEDULE WAS SUBSTANTIALLY COMPLIED WITH.

THE STATED PURPOSE OF THE RFP IS TO IMPLEMENT PHASE II BY INITIATING SELECTION AND ACQUISITION OF STANDARD ELECTRONIC DATA PROCESSING EQUIPMENT (EDPE) WHICH WILL PROCESS STANDARD AND UNIQUE COMMAND BASE LEVEL DATA SYSTEMS AT SELECTED AIR FORCE ACTIVITIES WORLDWIDE; PROVIDE STATE-OF-THE- ART MODULARITY TO ACCOMMODATE DATA PROCESSING WORKLOADS OF VARIOUS SIZES; REPLACE PRESENT COMBINATIONS OF EDPE AT SPECIFIED LOCATIONS; AND PROVIDE INITIAL EDPE FOR BASE LEVEL AIR FORCE ACTIVITIES PRESENTLY PROCESSING DATA MANUALLY. IN GENERAL, THE RFP PROVIDES THAT THE DATA SYSTEMS TO BE IMPLEMENTED WILL REQUIRE HIGH-SPEED CENTRAL PROCESSING, IMMEDIATE ACCESS STORAGE, MAGNETIC TAPE, PUNCHED CARD (I/O), PRINTER OUTPUT, PRIORITY INTERRUPT FEATURE AND REMOTE INPUT/OUTPUT DEVICES. THE EQUIPMENT IS TO BE SUPPORTED BY AN EXTENSIVE AND EFFICIENT SOFTWARE SYSTEM. IN ADDITION TO SPECIFIED MANDATORY REQUIREMENTS, CERTAIN DESIRED FEATURES WHICH WILL RECEIVE CONSIDERATION IN THE EVALUATION ARE LISTED. THE COMPUTER SYSTEMS MUST BE CAPABLE OF PROCESSING FOUR LEVELS OF WORKLOAD AND, FOR EVALUATION PURPOSES, 135 SYSTEMS ARE SPECIFIED IN DESIGNATED LEVELS, A, B, C, AND D, WITH A BEING THE SMALLEST AND D THE LARGEST, BROKEN DOWN AS FOLLOWS:

NUMBER OF

WORKLOAD LEVEL SYSTEMS

A 30

B 60

C 25

D 20

TOTAL 135

AS STATED IN THE RFP, IT IS ANTICIPATED THAT THE EDPE SELECTED WILL SATISFY THE BASE LEVEL DATA PROCESSING REQUIREMENTS FOR AT LEAST 6 YEARS. THE EQUIPMENT CONFIGURATIONS SELECTED FOR PURCHASE OR LEASE UNDER THE RFP ARE TO BE ORDERED FROM THE GSA FEDERAL SUPPLY SCHEDULE.

COST DATA TO BE SUBMITTED SEPARATE FROM THE SYSTEMS PROPOSAL AND TECHNICAL DATA WAS REQUIRED TO INCLUDE LEASE COSTS, LEASE WITH OPTION TO PURCHASE COSTS, PURCHASE COSTS, AND MAINTENANCE COSTS. IN ADDITION, THE RFP WAS AMENDED TO REQUIRE GUARANTEE OF PRICES QUOTED AT LEAST THROUGH JUNE 30, 1969.

CERTAIN PROVISIONS OF THE VOLUMINOUS RFP RELEVANT TO HONEYWELL'S PROTEST ARE QUOTED BELOW: ATTACHMENT I CHAPTER 1

* * * * * * * 1-16 REJECTION OF PROPOSALS:

1-16-1 THE GOVERNMENT RESERVES THE RIGHT TO REJECT ANY AND ALL PROPOSALS RECEIVED BY REASON OF THIS REQUEST, OR TO NEGOTIATE SEPARATELY WITH ANY SOURCE WHATSOEVER IN ANY MANNER NECESSARY TO SERVE THE BEST INTERESTS OF THE GOVERNMENT. THE UNITED STATES GOVERNMENT DOES NOT INTEND TO AWARD A CONTRACT SOLELY ON THE BASIS OF THIS REQUEST FOR PROPOSAL OR OTHER WISE PAY FOR THE INFORMATION SOLICITED OR OBTAINED. THE INFORMATION OBTAINED WILL BE USED IN DETERMINING THE SUITABILITY OF EQUIPMENT, PROCUREMENT OF WHICH WILL BE IN ACCORDANCE WITH GSA SCHEDULES. CHAPTER

* * ** * * * 2-3 EVALUATION CRITERIA:

2-3-1 GENERAL. ALL PROPOSALS RECEIVED WILL BE VALIDATED AND EVALUATED FOR COMPLIANCE WITH THE MANDATORY REQUIREMENTS OF THIS REQUEST FOR PROPOSAL. EQUIPMENT PROPOSALS WHICH DO NOT MEET THE MANDATORY REQUIREMENTS AND PROCESSING SPECIFICATIONS PRESCRIBED IN CHAPTER 4, ATTACHMENT II WILL NOT BE CONSIDERED FOR SELECTION. RESPONSIVE VENDOR PROPOSALS WILL BE EVALUATED BASED ON SYSTEMS PERFORMANCE, TECHNICAL CHARACTERISTICS, VENDOR SUPPORT, AND OVERALL COST TO THE AIR FORCE. ATTACHMENT 1 CHAPTER 2 (CONT-D)

2-3-2 LIVE TEST DEMONSTRATION. VENDORS WILL BE REQUIRED TO CONDUCT A LIVE TEST DEMONSTRATION OF THE EQUIPMENT/S) PROPOSED FOR WORKLOAD LEVEL D, USING THE SOFTWARE PROPOSED. (REFERENCE PARAGRAPH 1-2-3, CHAPTER 1, ATTACHMENT III). TIMINGS GENERATED BY THIS EXERCISE WILL BE USED TO ASSIST IN THE VALIDATION OF THE VENDOR'S TIMING TABLES.

2-4 MANDATORY REQUIREMENTS. VENDORS MUST COMPLY WITH THE FOLLOWING SUBPARAGRAPHS AS WELL AS THOSE MANDATORY REQUIREMENTS EXPRESSED IN ATTACHMENT II, CHAPTER 5. A STATEMENT INDICATING COMPLIANCE WITH EACH MANDATORY REQUIREMENT WILL BE INCLUDED IN THE VENDOR'S PROPOSAL, PART I, SECTION 2, ENTITLED "SPECIFIC REQUIREMENTS.' ALL COSTS ASSOCIATED WITH ANY OR ALL OF THE MANDATORY REQUIREMENTS MUST BE SHOWN IN PART III OF THE VENDOR'S PROPOSAL.

* * * * * * * CHAPTER 3

* * * * * * * SECTION 2. SPECIFIC REQUIREMENTS AND FEATURES:

2-1 MANDATORY REQUIREMENTS. THE VENDOR MUST PROVIDE IN THIS SECTION OF HIS PROPOSAL, A STATEMENT CONCERNING HIS POSITION WITH RESPECT TO EACH MANDATORY REQUIREMENT STATED IN PARAGRAPH 2-4, CHAPTER 2, ATTACHMENT I AND PARAGRAPH 5-2, CHAPTER 5, ATTACHMENT II OF THIS RFP. ATTACHMENT II CHAPTER 2

2-5-3 ONE OF THE MOST IMPORTANT REQUIREMENTS OF EDPE FOR PHASE II IS A HIGH DEGREE OF MODULARITY SO THAT THE EDPE CAPABILITY TO BE INSTALLED AT ANY LOCATION IS COMMENSURATE WITH THE WORKLOAD REQUIREMENTS. ATTACHMENT II CHAPTER 2 (CONT-D)

THROUGH THIS MODULARITY CONCEPT (WHICH WILL PERMIT FIELD EXPANSION OR CONTRACTION OF THE ACQUIRED EDPE AS REQUIRED TO MEET CHANGES IN WORKLOAD) IT IS EXPECTED THAT THE AIR FORCE BASE LEVEL DATA PROCESSING REQUIREMENTS CAN BE SATISFIED FOR AT LEAST A SIX YEAR PERIOD. CHAPTER 4

4-3 PROCESSING AND WORKLOAD STATEMENTS. THE WORKLOAD HAS BEEN IDENTIFIED AS HAVING THREE TYPES OF PROCESSING:

4-3-1 REAL TIME PROCESSING. TRANSACTIONS WHICH WILL BE INTRODUCED THROUGH REMOTE DEVICES AS THEY OCCUR DURING A PERIOD OF EIGHT HOURS A DAY, FIVE DAYS A WEEK.

4-3-2 BATCH PROCESSING. (CONCURRENT). THOSE APPLICATIONS WHICH MAY BE PROCESSED DURING THE EIGHT HOUR REAL TIME PERIOD OR OUTSIDE THE REAL TIME PERIOD AND WHICH ARE INDEPENDENT OF REAL TIME ACTIVITY.

4-3-3 BATCH PROCESSING. (NON-CONCURRENT). THOSE APPLICATIONS WHICH MUST BE PROCESSED OUTSIDE THE EIGHT HOUR REAL TIME PERIOD; THAT IS, DATA USED ARE DEPENDENT ON THE STATUS OF SELECTED DATA FILES AT THE CLOSE OF THE REAL TIME PERIOD.

4-7 LIVE TEST DEMONSTRATION. A LIVE TEST DEMONSTRATION IS REQUIRED FOR THIS PROJECT. ATTACHMENT III CONTAINS DETAILED DESCRIPTIONS OF THE APPLICATIONS WHICH ARE TO BE DEMONSTRATED.

CHAPTER 5

SPECIFIC CONDITIONS

5-1 GENERAL. THIS CHAPTER CONTAINS SPECIFIC AIR FORCE CONDITIONS AND REQUIREMENTS. MANDATORY REQUIREMENTS AND DESIRABLE FEATURES ARE IDENTIFIED. MANDATORY REQUIREMENTS MUST BE MET TO THE DEGREE STATED IN THESE SPECIFICATIONS IN ORDER FOR THE PROPOSALS TO RECEIVE CONSIDERATION. DESIRABLE FEATURES ARE NOT MANDATORY; HOWEVER, THEY DO REPRESENT ENHANCEMENTS OF THE SYSTEM WHICH WILL RECEIVE SPECIAL CONSIDERATION DURING THE EVALUATION OF PROPOSALS. EQUIPMENT PROPOSALS WHICH FAIL TO MEET ANY ONE OF THE MANDATORY REQUIREMENTS LISTED IN PARAGRAPH 2-4, CHAPTER 2, ATTACHMENT I AND THOSE BELOW WILL BE CONSIDERED NON-RESPONSIVE TO THIS REQUEST FOR PROPOSAL AND WILL NOT BE CONSIDERED FOR SELECTION.

5-2 MANDATORY REQUIREMENTS. THE FOLLOWING PARAGRAPHS PRESCRIBE MANDATORY REQUIREMENTS TO BE COMPLIED WITH IN RESPONDING TO THIS RFP. A STATEMENT CONCERNING THE VENDOR'S ABILITY TO FULFILL EACH MANDATORY REQUIREMENT LISTED MUST BE MADE IN PART I, SECTION 2, ENTITLED, "MANDATORY REQUIREMENTS," OF THE PROPOSAL. ALL COSTS ASSOCIATED WITH ANY REQUIREMENT LISTED BELOW WILL BE REFLECTED IN PART III OF THE PROPOSAL.

5-2-1 SYSTEM PERFORMANCE:

A. THE SYSTEMS PROPOSED MUST HAVE SUFFICIENT SPEED AND CAPACITY TO COMPLETELY PROCESS IN 200 HOURS OPERATIONAL USE TIME (AS DEFINED IN THE INSTRUCTIONS FOR TABLE B, EXHIBIT 4, ATTACHMENT I), EACH OF THE FOUR LEVELS OF WORKLOAD SHOWN IN CHAPTER 4, ATTACHMENT II.

B. AS A PART OF A LIVE TEST DEMONSTRATION, SELECTED REMOTE ACTIVITY WILL BE TIMED. THE FOLLOWING REMOTE TRANSACTIONS WHICH ARE SPECIFICALLY IDENTIFIED IN ATTACHMENT III, MUST BE PROCESSED WITHIN THE MAXIMUM RESPONSE INDICATED:

MAXIMUM

RESPONSE

TRANSACTION TIME ROUTINE 1510--- IFX STATUS OF ALLOTMENT 25 SECONDS IMMEDIATE INQUIRY, ITEM 106, EXHIBIT 3H 25 SECONDS

THE TIMING WILL INCLUDE TRANSMISSION OF DATA TO AND FROM THE COMPUTER AND WILL BEGIN WHEN THE LAST CHARACTER OF THE INPUT HAS BEEN ENTERED AT THE REMOTE UNIT AND WILL END WHEN THE LAST CHARACTER HAS BEEN PRINTED AT THE REMOTE UNIT. (REFERENCE STEPS 2A (1) AND (2), PARAGRAPH 2-3-2, CHAPTER 3, ATTACHMENT III).

5-2-2 AVAILABILITY OF EQUIPMENT:

A. ALL COMPONENTS OF EDPE PROPOSED IN RESPONSE TO THIS RFP, INCLUDING THOSE REQUIRED TO MEET EXPANSION REQUIREMENTS, MUST HAVE BEEN FORMALLY ANNOUNCED FOR MARKET PURPOSES ON OR BEFORE THE CLOSING DATE OF THIS RFP.

B. THE EQUIPMENT USED FOR THE LIVE TEST DEMONSTRATION MUST BE IDENTICAL TO THAT PROPOSED FOR THE "D" LEVEL INSTALLATION WITH THE EXCEPTIONS AS AUTHORIZED IN PARAGRAPH 1-2-3, CHAPTER 1, ATTACHMENT III.

PROPOSALS WERE SUBMITTED BY BURROUGHS, HONEYWELL, IBM, AND RCA. THE RFP PERMITTED ALTERNATE PROPOSALS AND IBM SUBMITTED THREE. HONEYWELL SUBMITTED ONE. THE PROPOSALS WERE EVALUATED IN ACCORDANCE WITH A SELECTION PLAN PREPARED BY THE SOURCE SELECTION ADVISORY COUNCIL (SSAC) PRIOR TO ISSUANCE OF THE RFP. FIVE MAJOR WORKING GROUPS COMPOSED OF PERSONNEL FROM VARIOUS AIR FORCE ACTIVITIES PERFORMED THE EVALUATION TASKS, COORDINATING THEIR FUNCTIONS WITH AND REPORTING THEIR FINDINGS TO THE SOURCE SELECTION EVALUATION BOARD (SSEB) WHICH, IN TURN, REPORTED TO THE SSAC. THE SSAC'S RECOMMENDATION WAS SUBMITTED TO THE SOURCE SELECTION AUTHORITY (SSA), ASSISTANT SECRETARY MARKS, FOR APPROVAL.

THE FIVE MAJOR WORKING GROUPS AND THEIR AREAS OF RESPONSIBILITY IN THE EVALUATION PROCESS WERE:

(1) TECHNICAL CHARACTERISTICS GROUP: THIS GROUP VALIDATED VENDOR'S COMPLIANCE WITH MANDATORY REQUIREMENTS AND EVALUATED TECHNICAL CHARACTERISTICS OF EQUIPMENT PROPOSED. (2) VENDOR SUPPORT GROUP: THIS GROUP VALIDATED PROGRAM TEST TIME, COMPLETENESS OF MANUALS AND TECHNICAL LITERATURE (DOCUMENTATION) AND MAINTENANCE SUPPORT.

(3) SOFTWARE GROUP: THIS GROUP DETERMINED THE RESPONSIVENESS OF VENDORS PROPOSED SOFTWARE TO THE MANDATORY SOFTWARE REQUIREMENTS. A COMPARATIVE MEASUREMENT OF THE PERFORMANCE OF EACH WAS VALIDATED THROUGH EXTENSIVE ANALYSIS OF THE LITERATURE PROVIDED BY THE VENDORS AND ON THE SOFTWARE OBSERVED DURING THE LIVE TEST DEMONSTRATION.

(4) SYSTEMS PERFORMANCE GROUP: THIS GROUP REVIEWED, VALIDATED AND/OR ADJUSTED THE TIMING SUBMITTED IN VENDORS' PROPOSALS. THE SYSTEMS PERFORMANCE WORK GROUP PERFORMED TIMING FUNCTIONS AS MEMBERS OF THE LIVE TEST DEMONSTRATION TEAM. THE TIMINGS DERIVED DURING THE LIVE TEST DEMONSTRATION WERE UTILIZED IN THE VALIDATION AND/OR ADJUSTMENTS OF VENDORS SUBMITTED TIMES.

(5) COST GROUP: THIS GROUP REVIEWED THE COST PROPOSALS TO INSURE COMPLIANCE WITH THE APPROPRIATE INSTRUCTIONS CONTAINED IN THE RFP, VALIDATED THE COST COMPLIATIONS SUBMITTED BY VENDORS, AND DEVELOPED DETAILED COST ANALYSIS IN ACCORDANCE WITH THE COST EVALUATION PLAN. COST DATA WERE NOT MADE AVAILABLE TO MEMBERS OF OTHER EVALUATION WORK GROUPS. HOWEVER, OTHER GROUPS PROVIDED DATA TO BE USED IN CALCULATING COST AS REQUIRED FROM THE COMPUTED OPERATIONAL USE TIMES.

AS A RESULT OF EVALUATION IN ACCORDANCE WITH THE FOREGOING PROCEDURE IT WAS DETERMINED THAT ONLY IBM MET ALL OF THE MANDATORY REQUIREMENTS OF THE RFP. IN FACT, ALL THREE OF IBM'S PROPOSALS ARE REPORTED RESPONSIVE TO THE RFP. IT IS REPORTED THAT BURROUGHS, RCA, AND HONEYWELL WERE UNABLE TO DEMONSTRATE IN THE LIVE TEST THE SPEED AND CAPACITY TO PROCESS WORKLOAD LEVELS C AND D IN 200 HOURS OPERATIONAL USE TIME AS REQUIRED BY PARAGRAPH 5-2-1, CHAPTER 5, ATTACHMENT II. BURROUGHS AND RCA ARE ALSO REPORTED DEFICIENT IN OTHER AREAS. WITH REGARD TO HONEYWELL'S PROPOSAL, IT IS REPORTED THAT THE INITIAL EXAMINATION INDICATED WELL-BALANCED SYSTEMS HAD BEEN PROPOSED FOR EACH WORKLOAD LEVEL AND THAT FROM THE DESK CHECK OF THE PROPOSAL MADE PRIOR TO THE BENCHMARK IT APPEARED THAT THE HONEYWELL EQUIPMENT HAD ADEQUATE MEMORY, THE REQUIRED INTERNAL SPEEDS AND AMPLE PERIPHERALS. HOWEVER, THE LIVE TEST DEMONSTRATION, USING TEST DATA IDENTICAL, EXCEPT FOR VOLUME AND CONTENT, TO THE SAMPLE TEST DATA PROVIDED PRIOR TO SUBMISSION OF PROPOSALS, SHOWED A CUMBERSOME TECHNIQUE FOR HANDLING THE REAL-TIME PROCESSING. AS A RESULT, OPERATIONAL USE HOURS OF 266.7 AND 260.8 WERE DETERMINED FOR WORKLOAD LEVELS C AND D, RESPECTIVELY.

SINCE ONLY IBM WAS FOUND RESPONSIVE TO ALL MANDATORY REQUIREMENTS, THE OTHER OFFERORS WERE NOT CONSIDERED FURTHER. THEREFORE, A COST ANALYSIS AND WEIGHING OF THE THREE IBM PROPOSALS WAS PERFORMED AND THE SSAC RECOMMENDED SELECTION OF THE PROPOSAL WHICH WAS DETERMINED TO HAVE A $146,544,354 PURCHASE COST BASED ON THE 6-YEAR LIFE STATED IN THE RFP. BEFORE FINAL SELECTION WAS MADE A "COST EFFECTIVENESS STUDY" WAS CONDUCTED TO DETERMINE WHETHER THE COST OF THE IBM PROPOSAL WAS REASONABLE. THIS STUDY TOOK THE FORM OF A COMPARISON OF THE TOTAL SYSTEMS COST FOR 6 YEARS BETWEEN THE HONEYWELL AND IBM PROPOSALS. IT WAS CONCLUDED THAT ALTHOUGH THE INITIAL COST OF THE IBM PROPOSAL WAS APPROXIMATELY $65-70 MILLION MORE THAN HONEYWELL-S, THE OVERALL COSTS FOR 6 YEARS ARE APPROXIMATELY EQUAL. HOWEVER, THE STUDY ALSO INDICATES THAT BECAUSE OF THE HONEYWELL'S RELATIVE INEFFICIENCY IN HANDLING REAL-TIME APPLICATIONS AND BECAUSE OF CONSERVATIVE WORKLOAD GROWTH RATE FIGURES USED IN THE STUDY, THE IBM PROPOSAL IS DECIDEDLY MORE ADVANTAGEOUS. ON APRIL 12, 1967, THE SSA ANNOUNCED SELECTION OF IBM.

WE UNDERSTAND THAT ALTHOUGH NO ORDERS HAVE BEEN PLACED WITH IBM FOR DELIVERY OF EQUIPMENT, TRAINING OF AIR FORCE PERSONNEL BY IBM HAS BEGUN PURSUANT TO THE TERMS OF THE RFP; FURTHER, THAT TWO SYSTEMS WILL BE INSTALLED THE LATTER PART OF JULY FOR TESTING AT IBM'S EXPENSE, AND THAT THE FIRST ORDER IS TO BE PLACED WITH IBM AT ABOUT THE SAME TIME FOR DELIVERY OF TWO SYSTEMS IN AUGUST. INSTALLATION THEREAFTER WILL PROCEED AS OUTLINED IN THE RFP.

THE PROTEST TO OUR OFFICE WAS INITIATED BY A TELEGRAM FROM HONEYWELL DATED MAY 11, 1967. THE SPECIFIC BASIS OF THE PROTEST WAS SET FORTH AT SOME LENGTH IN A LETTER DATED MAY 22, 1967, FROM THE ATTORNEYS REPRESENTING HONEYWELL. THIS HAS BEEN SUPPLEMENTED BY ADDITIONAL CORRESPONDENCE. THE BASIC ARGUMENT ADVANCED BY HONEYWELL IS THAT THE AIR FORCE HAS REFUSED TO CONDUCT ORAL OR WRITTEN DISCUSSIONS WITH IT CONCERNING THE ALLEGED TECHNICAL DEFICIENCY OR DEFICIENCIES IN ITS PROPOSAL AND THAT SUCH REFUSAL IS CONTRARY TO THE MANDATE OF 10 U.S.C. 2304 (G) WHICH PROVIDES, IN SUBSTANCE, THAT IN ALL NEGOTIATED PROCUREMENTS IN EXCESS OF $2,500 WRITTEN OR ORAL DISCUSSIONS SHALL BE CONDUCTED WITH ALL RESPONSIBLE OFFERORS WHO SUBMIT PROPOSALS WITHIN A COMPETITIVE RANGE, PRICE, AND OTHER FACTORS CONSIDERED.

IT IS HONEYWELL'S POSITION THAT THIS STATUTE ESTABLISHED A REQUIREMENT THAT THERE BE DISCUSSIONS WITH ALL RESPONSIBLE OFFERORS WITHIN A COMPETITIVE RANGE REGARDLESS OF RESPONSIVENESS; AND THAT SINCE IT IS RESPONSIBLE AND WITHIN A COMPETITIVE RANGE, APPROXIMATELY $60,000,000 BELOW IBM, THERE IS A DUTY TO DISCUSS ANY ALLEGED TECHNICAL DEFICIENCIES. THESE DISCUSSIONS WILL REVEAL, HONEYWELL SAYS, THAT MINOR CORRECTIONS WILL CURE ANY DEFICIENCIES AND THAT THE COST OF SUCH CORRECTIONS WILL BE MINIMAL, $1,250,000, IN RELATION TO THE $60,000,000 DISPARITY BETWEEN ITS PROPOSAL AND IBM'S PROPOSAL. IN THIS CONNECTION, HONEYWELL HAS REFERRED TO AN EXCHANGE OF LETTERS WITH YOU. THEY HAVE ATTACHED COPIES OF THEIR LETTER OF MAY 1 TO YOU, AS WELL AS YOUR REPLY OF MAY 5. THE FOLLOWING EXCERPTS FROM HONEYWELL'S LETTER INDICATE ITS POSITION WITH REGARD TO THE NATURE OF ANY DEFICIENCIES AND THE NECESSITY FOR AND BENEFIT TO BE DERIVED FROM SUCH DISCUSSIONS: BECAUSE THE ACTUAL WORKLOAD ITSELF COULD NOT BE RUN IN THE LIVE TEST DEMONSTRATION, IT WAS NECESSARY TO "EXTRAPOLATE" (A WORD USED BY THE DEBRIEFING COMMITTEE) OUR TIMINGS FROM DATA CONTAINED IN OUR MANUALS AND SIMILAR OPERATIONS PERFORMED DURING THE LIVE TEST DEMONSTRATION. THIS EXTRAPOLATION IS A VERY TECHNICAL STUDY, AND ONE IN WHICH TECHNICIANS MAY EASILY DIFFER IN REACHING CONCLUSIONS. IN FACT, OUR TECHNICAL PEOPLE CONSIDERED THAT WE HAD MET ALL THE TIMING REQUIREMENTS AND WERE ASTOUNDED AT THE RESULT REACHED BY THE EVALUATION COMMITTEE. HAD WE NOT BEEN SURE FROM THE OUTSET THAT WE COULD PERFORM THE REQUESTED WORKLOAD, WE WOULD NOT HAVE SPENT $1,500,000 IN OUR PROPOSAL EFFORT. THE DEBRIEFING WE ASKED FOR MORE DETAIL CONCERNING THE ACTUAL EXTRAPOLATION PROCESS SO THAT WE COULD RECONCILE THE EVALUATION COMMITTEE RESULTS WITH OUR OWN. WE ASKED FOR MORE DETAIL CONCERNING THE COST EFFECTIVENESS STUDY. TO BOTH OF THESE QUESTIONS WE WERE TOLD THE SOURCE SELECTION EVALUATION BOARD DID NOT HAVE AUTHORITY TO REVEAL THIS INFORMATION. IN PREVIOUS AIR FORCE AND OTHER GOVERNMENT EDP PROCUREMENTS IN WHICH WE HAVE BEEN INVOLVED, AND WHERE THERE WERE DIFFERENCES IN THE TIMING EVALUATION WITH THOSE PROPOSED, THE GOVERNMENT HAS REGULARLY RECONCILED WITH US THOSE PROPOSED WITH THE EVALUATION. WE REPEAT, THAT WHERE THE WORKLOAD IS NOT ACTUALLY PERFORMED BY A LIVE TEST DEMONSTRATION, THE DETERMINATION OF TIMINGS, PARTICULARLY IN A MULTI-PROCESSING MODE OF OPERATION, IS A VERY TECHNICAL CONSIDERATION--- EASILY GIVEN TO MISUNDERSTANDINGS OF EQUIPMENT SPECIFICATION ON THE ONE HAND AND WORKLOAD ON THE OTHER. A RECONCILIATION PROCEDURE WOULD BE NORMAL IN SUCH A SITUATION. WE ASKED THE DEBRIEFING COMMITTEE WHY THIS HAD NOT BEEN DONE IN THIS PROCUREMENT. WE WERE TOLD THAT THE AIR FORCE PROCUREMENT REGULATIONS DO NOT REQUIRE IT, IN THIS TYPE OF PROCUREMENT, AND THAT, IN ANY CASE, THERE WAS NOT SUFFICIENT TIME. THE TECHNICAL PROPOSAL WAS SUBMITTED FOUR MONTHS BEFORE NOTICE OF INTENDED AWARD. ONLY FOUR FIRMS PROPOSED. OUR INFORMATION IS THAT ONLY ONE BIDDER MET ALL THE MANDATORY REQUIREMENTS. * * * IN THE DEBRIEFING, THE AIR FORCE RECOMMENDED THAT WE REEXAMINE OUR COMMUNICATIONS HANDLING. WE HAVE HAD THREE TEAMS, INCLUDING A CONSULTANT, EXAMINE THIS AREA AND REEVALUATE OUR WORKLOAD ESTIMATES. HAVE FOUND AN OMISSION IN OUR ORIGINAL ESTIMATES AND WE HAVE CONCLUDED THAT SYSTEM OPERATION OUTLINED IN OUR PROPOSAL IS LIKELY TO EXCEED THE 200 HOUR LIMIT BY A SMALL AMOUNT. WE FOUND NO SUPPORT FOR THE STATEMENT THAT WE SUBSTANTIALLY EXCEEDED THE ALLOWED TIME. THE OMISSION IS AN OBVIOUS CONTRADICTION WITH OTHER PORTIONS OF OUR PROPOSAL. HAD THE AIR FORCE ATTEMPTED TO CLARIFY THE CONTRADICTION, THE OVERAGE COULD HAVE BEEN EASILY REMEDIED BY ADDITION OF MEMORY AND BY SUBSTITUTION OF A FEW SOFTWARE INSTRUCTIONS. THESE CORRECTIONS COULD HAVE BEEN ACHIEVED AT A PRICE WHICH WOULD NOT EXCEED $1,250,000, AND WITH NO DELAY IN DELIVERY. RESPECTFULLY, BUT URGENTLY, REQUEST THAT YOU REQUIRE THE AIR FORCE SYSTEMS COMMAND TO EXPLAIN THE METHOD OF ARRIVING AT THEIR TIMING CONCLUSIONS TO US, AND GIVE US AN OPPORTUNITY TO RECONCILE THESE TIMINGS WITH OUR OWN CONCLUSIONS. * * *

HONEYWELL'S ATTORNEYS POINT TO YOUR REPLY AS INDICATING A MISCONCEPTION AS TO THE OBLIGATION OF THE AIR FORCE UNDER THE CITED STATUTE AND APPLICABLE REGULATIONS TO CONDUCT SUCH DISCUSSIONS. YOUR LETTER OF MAY 5, 1967, TO HONEYWELL IS QUOTED IN FULL BELOW:

THIS IS IN RESPONSE TO YOUR LETTER OF MAY 1, 1967, AND THAT OF MR. SPANGLE ALSO DATED MAY 1, REGARDING THE RECENT SELECTION OF A CONTRACTOR FOR THE USAF BASE LEVEL STANDARDIZATION PROGRAM. BOTH LETTERS POINT UP THE DEFICIENCY REGARDING THE MANDATORY 200-HOUR OPERATIONAL USE TIME REQUIREMENT FOR WORKLOAD LEVELS C AND D. AS NOTED IN THE DEBRIEFING CONDUCTED AT L. G. HANSCOM FIELD AND ACKNOWLEDGED IN YOUR LETTER, THESE CONFIGURATIONS EXCEEDED THE LIMIT. THE LIVE TEST DEMONSTRATION TIMES CONTAINED IN THE ATTACHMENT TO YOUR LETTER ARE ESSENTIALLY THE SAME AS THOSE RECORDED BY THE AIR FORCE. THE VARIANCES WOULD NOT BE SUCH AS TO CAUSE DIFFERENT CONCLUSIONS REGARDING THE ABILITY OF HONEYWELL CONFIGURATIONS C AND D TO MEET THE MANDATORY 200-HOUR REQUIREMENT.

BECAUSE OF THE IMPORTANCE OF THIS MEASURE TO THE ULTIMATE DECISION, WE REVIEWED THE 200-HOUR REQUIREMENT AND VALIDATED IT AS AN APPROPRIATE MEANS TO JUDGE THE CAPABILITY OF THE EQUIPMENT AND SOFTWARE PROPOSED TO ACCOMPLISH THE ANTICIPATED AIR FORCE WORKLOAD REQUIREMENTS.

THE POINT IS MADE IN MR. SPANGLE'S LETTER THAT THE SSEB DID NOT ATTEMPT TO CONSIDER THE COST OF AUGMENTING YOUR SYSTEM WITH ADDITIONAL EQUIPMENT NOT INCLUDED IN YOUR PROPOSAL. THIS IS CORRECT. BECAUSE THE SYSTEM REQUIREMENTS WERE TIGHTLY DRAWN AND THE BENCHMARK TEST DEVELOPED TO MEASURE AGAINST THOSE REQUIREMENTS, THE SSEB WAS NOT EXPECTED TO SPECULATE WITH REGARD TO POTENTIAL PERFORMANCE OF CONTRACTORS' SYSTEMS NOT SUBMITTED IN ACCORDANCE WITH THE RULES GOVERNING THE RFP. TO HAVE DONE SO, OR TO HAVE PERMITTED ANY CONTRACTOR TO MAKE AMENDMENTS AFTER SUBMITTING A PROPOSAL IS PROHIBITED BY ASPR, AND WOULD BE PATENTLY UNFAIR TO OTHER COMPETITORS.

CERTAIN POTENTIAL COMPETITORS ELECTED NOT TO SUBMIT A BID WELL IN ADVANCE OF THE SUBMISSION DATE BECAUSE THEY CONCLUDED THEY DID NOT HAVE SUFFICIENT TIME TO ACCOMPLISH THE NECESSARY HARDWARE/SOFTWARE COMBINATION TO MEET OUR DEMANDING SPECIFICATIONS. YOU CAN IMAGINE THE REACTION OF THOSE FIRMS, AS WELL AS THE OTHER ACTIVE COMPETITORS, WERE WE TO PERMIT AMENDMENTS TO ANY SINGLE CONTRACTOR'S PROPOSAL AFTER SUBMISSION. IF WE WERE INSTEAD TO TAKE THE COURSE OF REOPENING THE COMPETITION, WE SHOULD NOT BE ABLE TO ACCOMPLISH PROCUREMENT IN TIME TO MEET THE URGENT DATA PROCESSING REQUIREMENTS OF OUR FUNCTIONAL MANAGERS.

THE TRANSCRIPT OF THE DEBRIEFING INDICATES THAT YOUR DEVELOPMENT TEAM MAY BELIEVE THAT HONEYWELL COULD HAVE SUBMITTED A COMBINATION OF EQUIPMENT AND SOFTWARE WHICH MET THE MANDATORY REQUIREMENTS AND WOULD HAVE BEEN LESS EXPENSIVE THAN PROPOSALS BY OTHER BIDDERS. BUT, I THINK WE AGREE YOU DID NOT. A MULTIPLE STEP TECHNICAL EVALUATION ALLOWING FOR EXTENDED NEGOTIATION AND CORRECTION FOLLOWED BY PRICE COMPETITION AND SELECTION WHICH COULD POSSIBLY ANSWER THAT QUESTION HAS IN OUR EXPERIENCE THE CHARACTER OF A PAPER COMPETITION WITH QUESTIONS ARISING WHETHER THE GOVERNMENT IS DEALING ON EQUAL TERMS WITH ALL COMPETITORS. IT PROVIDES PROMISES OF TECHNICAL ACCOMPLISHMENT AND PERFORMANCE RATHER THAN DEMONSTRABLE EVIDENCE THAT CONTRACTUAL REQUIREMENTS ARE FULLY UNDERSTOOD AND CAN BE MET. OUR EXPERIENCE IN THIS TYPE OF COMPETITION IS OFTEN DISAPPOINTING, BOTH IN THE PRODUCT AND SERVICE PROVIDED AND, THE ULTIMATE PRICE PAID. IT IS FOR THESE REASONS THAT FOR THIS COMPETITION WE SELECTED THE BENCHMARK APPROACH, WITH THE REQUEST FOR PROPOSALS STIPULATING BOTH MANDATORY REQUIREMENTS AND A DEFINITIVE TIME LIMITATION FOR MEETING THESE REQUIREMENTS IN THE COMPETITION.

THERE WAS A COST EFFECTIVENESS ANALYSIS PERFORMED AT THIS HEADQUARTERS AFTER RECEIPT OF THE SSAC RECOMMENDATION. THIS ANALYSIS WHICH CONFIRMED THE IMPORTANCE OF THE 200-HOUR OPERATIONAL USE AS A MANDATORY REQUIREMENT ALSO INDICATED THAT THE SELECTION OF THE IBM EQUIPMENT IS COST EFFECTIVE.

MR. SPANGLE EXPRESSED CONCERN IN HIS LETTER THAT, FROM THE LETTER OF FEBRUARY 10TH, WE MIGHT HAVE CONCLUDED THE OPERATING SYSTEM DEFICIENCIES MENTIONED THEREIN CONTRIBUTED TO YOUR FIRM'S EXCEEDING THE 200-HOUR REQUIREMENT. MY TECHNICAL ADVISORS TELL ME THEY COULD NOT IDENTIFY ALL THE FACTORS WHICH CAUSED YOUR EXCEEDING 200 HOURS. THEY ALLOW, THEREFORE, THAT THE DEFICIENCIES NOTED IN THE FEBRUARY 10TH LETTER MAY NOT HAVE BEEN SPECIFIC CONTRIBUTING FACTORS TO THE FAILURE TO SATISFY THE BENCHMARK REQUIREMENT. WHATEVER THE CAUSES, WE BOTH RECOGNIZE THAT THE 200-HOUR MANDATORY REQUIREMENT WAS EXCEEDED. IN THIS PROJECT AS IN PREVIOUS AIR FORCE SELECTION PROJECTS, INCLUDING THE MAJOR AIR COMMAND PROGRAM, IT HAS BEEN OUR POLICY TO DISCUSS ANY INDISTINCT ASPECTS OF A PROPOSAL DURING THE EVALUATION PROCESS. AIR FORCE COMMUNICATIONS WITH HONEYWELL, AND THE RESULTANT ADJUSTMENTS PERMITTED TO YOUR PROPOSAL SUBSEQUENT TO ITS SUBMISSION, WERE DISCUSSED IN THE FORMAL TEXT OF THE DEBRIEFING. THOSE ADJUSTMENTS WERE CONFINED TO CORRECTIONS OF OBVIOUS OMISSIONS OR MISUNDERSTANDINGS THAT COULD BE READILY IDENTIFIED AND CORRECTED. YOUR STATEMENT IN REGARD TO RECONCILIATION OF TIMINGS,"WE WERE TOLD THAT THE AIR FORCE PROCUREMENT REGULATIONS DO NOT REQUIRE IT, IN THIS TYPE OF PROCUREMENT, AND THAT, IN ANY CASE, THERE WAS NOT SUFFICIENT TIME," CAUSES ME SOME CONCERN. LET ME STATE, WITHOUT EQUIVOCATION, THAT TIME DURING THE EVALUATION PERIOD WAS NOT A FACTOR IN THE CONSIDERATION OF WHETHER THE AIR FORCE SHOULD OR SHOULD NOT DISCUSS WITH HONEYWELL THE NONRESPONSIVENESS TO THIS 200-HOUR REQUIREMENT. THE REFERENCE TO TIME IN THE DEBRIEFING WAS RELATED TO WHEN AND HOW NON-RESPONSIVE PROPOSERS WOULD BE NOTIFIED. WAS INDICATED THAT THERE WAS INSUFFICIENT TIME TO NOTIFY LOSING PROPOSERS AFTER APPROVAL OF THE RECOMMENDATIONS AND PRIOR TO ANNOUNCEMENT OF THE SELECTED PROPOSER.

LATE PROPOSALS AND AMENDMENTS ARE GOVERNED BY ARMED SERVICES PROCUREMENT REGULATIONS. THESE DIRECTIVES PROHIBIT CONSIDERATION OF MODIFICATIONS AFTER THE TIME SPECIFIED FOR SUBMISSION OF THE PROPOSAL (WITH CERTAIN EXCEPTIONS NOT APPLICABLE HERE). IN THIS LIGHT, THE MODIFICATION OF A PROPOSAL TO INCLUDE CHANGES TO MAIN MEMORY, SOFTWARE TECHNIQUES, AND CHANGES IN TIMING TABLES WOULD BE CONSIDERED IN THE NATURE OF A LATE PROPOSAL.

HAVING REFLECTED ON OUR MAY 1 DISCUSSION AND HAVING REVIEWED CAREFULLY THE POINTS RAISED IN THE LETTERS FROM MR. SPANGLE AND YOURSELF, I AM FULLY SATISFIED THAT THE SELECTION OF IBM FOR THE AIR FORCE PHASE II PROCUREMENT RECOMMENDED BY THE SOURCE SELECTION ADVISORY COMMITTEE AND APPROVED IN THIS HEADQUARTERS WAS A CORRECT DECISION. THEREFORE, I MUST DENY YOUR REQUESTS FOR FURTHER CONSIDERATION AND FOR RESCINDING THE AWARD.

IN SUPPORT OF ITS INTERPRETATION OF 10 U.S.C. 2304 (G) AND APPLICABILITY THEREOF TO THE INSTANT CASE, HONEYWELL HAS CITED SEVERAL DECISIONS OF OUR OFFICE AND PROVISIONS OF THE ARMED SERVICES PROCUREMENT REGULATION (ASPR), WHICH WILL BE DISCUSSED HEREINAFTER.

IN THE ALTERNATIVE, HONEYWELL CONTENDS THAT ITS PROPOSAL WAS NOT IN FACT NONRESPONSIVE SINCE IT PROPOSED AND WOULD BE BOUND TO DELIVER A SYSTEM WHICH OPERATES WITHIN THE 200 HOURS. IN THIS CONNECTION, THE AIR FORCE'S FAILURE TO NOTIFY HONEYWELL THAT ITS PROPOSAL WAS "UNACCEPTABLE" PURSUANT TO THE PROVISIONS OF ASPR 3-508 IS CITED AS INDICATING THAT THE PROPOSAL WAS NOT IN FACT CONSIDERED NONRESPONSIVE, BUT WAS RULED OUT ON THE BASIS OF THE COST EFFECTIVENESS STUDY WHICH IS INCONSISTENT WITH A FINDING OF UNACCEPTABILITY OR NONRESPONSIVENESS.

HONEYWELL ALSO TAKES ISSUE WITH THE STATEMENT IN YOUR LETTER OF MAY 5, 1967, THAT MODIFICATION OF ITS PROPOSAL WOULD BE A LATE MODIFICATION AND NOT PROPERLY FOR CONSIDERATION IN VIEW OF THE PROVISIONS OF ASPR 3-506 (A). IT IS CONTENDED THAT MODIFICATION WOULD BE PROPER UNDER THE PROVISIONS OF ASPR 3-506 (H) WHICH PROVIDES:

(H) THE NORMAL REVISIONS OF PROPOSALS BY SELECTED OFFERORS OCCURRING DURING THE USUAL CONDUCT OF NEGOTIATIONS WITH SUCH OFFERORS ARE NOT TO BE CONSIDERED AS LATE PROPOSALS OR LATE MODIFICATIONS, BUT SHALL BE HANDLED IN ACCORDANCE WITH 3-805.1 (B).

HONEYWELL QUESTIONS THE DELETION OF A MANDATORY REQUIREMENT WHICH ALLEGEDLY BENEFITED ONLY IBM, AND ALSO QUESTIONS IBM'S COMPLIANCE WITH ANOTHER MANDATORY REQUIREMENT. THE FIRST CONTENTION INVOLVES THE CHANGE FROM MANDATORY TO DESIRABLE THAT PART OF PARAGRAPH 5-2-9H (5), ATTACHMENT II, WHICH REQUIRED AUTOMATIC RETRANSMISSION AS A MEANS OF ERROR CORRECTION. IT IS HONEYWELL'S CONTENTION THAT THIS CHANGE WAS MADE ON OCTOBER 5, 1966, BY ADDENDUM FOR THE PURPOSE OF PERMITTING IBM TO AVOID AN EXPENSIVE SUBCONTRACT FOR A CAPABILITY ITS EQUIPMENT DID NOT PROVIDE. A RESULT, HONEYWELL SAYS, ITS COMPETITIVE STANDING IN THE PRICE- PERFORMANCE TRADE-OFF WAS PREJUDICED BECAUSE IT WAS HEAVILY COMMITTED TO PROVIDING SUCH A FEATURE IN ITS EQUIPMENT WHEN THE DELETION WAS MADE. THE CONTENTION WITH RESPECT TO IBM'S RESPONSIVENESS INVOLVES THE MULTI- PROGRAMMING MODE OF OPERATION REQUIREMENT AT THE D LEVEL. IT IS ARGUED THAT IBM COULD NOT BE RESPONSIVE TO THIS REQUIREMENT BECAUSE ITS EQUIPMENT HAS ONLY FIVE MAGNETIC TAPE UNITS.

FINALLY, HONEYWELL TAKES THE POSITION THAT IF NEGOTIATIONS WITH IT PURSUANT TO 10 U.S.C. 2304 (G) ARE NOT CONSIDERED APPROPRIATE, THEN THE RFP SHOULD BE CANCELED AND THE REQUIREMENT RESOLICITED BECAUSE THERE HAS NOT BEEN EFFECTIVE COMPETITION AND THE ONLY PROPOSAL CONSIDERED IS UNREASONABLE.

THE AIR FORCE'S REPLY TO EACH OF THE FOREGOING ARGUMENTS IS INCLUDED IN ITS REPORT ON AIR FORCE PHASE II SELECTION, WHICH WAS TRANSMITTED ALONG WITH OTHER DOCUMENTS BY ASSISTANT SECRETARY MARKS' LETTER OF JUNE 16, 1967. THAT PART OF THE REPORT DEALING WITH EACH OF HONEYWELL'S CONTENTIONS IS QUOTED BELOW. III. THE GROUNDS OF PROTEST. ARGUMENT I

HONEYWELL CONTENDS THAT BEFORE REJECTING ITS PROPOSAL AS NON RESPONSIVE THE AIR FORCE WAS REQUIRED BY 10 U.S.C. SEC. 2304 (G) TO GIVE IT THE OPPORTUNITY TO "NEGOTIATE," THAT IS, TO AMEND ITS PROPOSAL. STRESSING THE DIFFERENCE IN INITIAL ACQUISITION COST THE PROTESTANT ARGUES THAT ITS PROPOSAL SHOULD HAVE BEEN CONSIDERED WITHIN A COMPETITIVE RANGE AND SHOULD NOT HAVE BEEN ELIMINATED FROM THE COMPETITION WITHOUT SUBSEQUENT NEGOTIATIONS. AIR FORCE RESPONSE

THE SIMPLE FACT IS THAT HONEYWELL'S PROPOSAL WAS NOT EASILY REPARABLE. IT EXCEEDED THE MANDATORY REQUIREMENTS NOT BY A SLIGHT BUT BY A SUBSTANTIAL MARGIN. IT WAS NOT WITHIN A COMPETITIVE RANGE. THUS, EVEN APPLYING THE STANDARD OF 10 U.S.C. SEC. 2304 (G), THERE WAS NO REQUIREMENT TO NEGOTIATE WITH HONEYWELL.

A MORE FUNDAMENTAL OBJECTION TO HONEYWELL'S ARGUMENT IS THAT THE REFERENCE TO NEGOTIATIONS IS MISLEADING IN THIS UNIQUE AREA OF ADPE ACQUISITION OF FEDERAL SUPPLY SCHEDULE CONTRACTS. NO SEPARATE CONTRACT WITH TERMS AND CONDITIONS AND PRICE DIFFERENT FROM THE GSA'S FEDERAL SUPPLY SCHEDULE ARE TO BE ENTERED INTO THROUGH NEGOTIATION BY THE AIR FORCE. ALTHOUGH THE AIR FORCE DETERMINES ITS SYSTEMS REQUIREMENTS AND SELECTS THE EQUIPMENT WHICH CAN BEST MEET THOSE REQUIREMENTS PURSUANT TO P.L. 89-306, IT IS REQUIRED TO ORDER THE EQUIPMENT UNDER THE GSA SCHEDULE CONTRACTS. HOLDERS OF THOSE CONTRACTS MAY, TO IMPROVE THEIR COMPETITIVE POSITION, OFFER PRICE REDUCTIONS. SUCH REDUCTIONS WERE IN FACT OFFERED BY SOME OF THE PHASE II, COMPETITORS. REDUCTIONS ARE CONTEMPLATED BY THE SCHEDULE CONTRACTS AND BY THEIR TERMS INURE TO OTHER GOVERNMENT AGENCIES ORDERING LIKE QUANTITIES UNDER LIKE CIRCUMSTANCES. (SEE B-148889, AUGUST 8, 1962) GENERALLY, DICKERING OVER PRICE BY THE ORDERING AGENCY IS NOT PART OF THE FEDERAL SUPPLY SCHEDULE SYSTEM, SINCE GSA NEGOTIATES THE PRICES AND TERMS OF SUCH CONTRACTS. THUS, IT IS DOUBTFUL WHETHER A SELECTION OF A SOURCE UNDER THE FEDERAL SUPPLY SCHEDULE SYSTEM IS A SITUATION WHERE NEGOTIATIONS ARE BEING CONDUCTED WITH ANY OF THE POSSIBLE SOURCES, WITHIN THE MEANING OF 10 U.S.C. SEC. 2304 (G). IN SUCH SELECTIONS, THE GOVERNMENT'S INTEREST IN PERFORMANCE AND PRICE ARE PROTECTED BY THE PROCESS OF COMPETITION. AS INDICATED ABOVE, SEVERAL COMPETITORS DID OFFER PRICES LOWER THAN THEIR GSA SCHEDULE LISTINGS.

HONEYWELL'S COMPLAINT IS REALLY NOT THAT NEGOTIATIONS WERE NOT CARRIED ON WITH THEM, BUT THAT THEY WERE NOT GIVEN ADDITIONAL TIME AFTER THE BENCHMARK TEST ESTABLISHED THAT THEIR PROPOSAL DID NOT MEET THE 200 OPERATIONAL USE HOURS REQUIREMENT, TO REVISE THEIR PROPOSAL, ADDING NEW EQUIPMENT, AND TO RESUBMIT IT FOR EVALUATION AND TEST BY THE AIR FORCE. THE AIR FORCE'S POSITION IS THAT UNDER THE CIRCUMSTANCES IT WAS NOT IMPROPER TO DENY HONEYWELL (AND THE OTHER LOSERS) THIS OPPORTUNITY.

THIS CONCLUSION RESTS ON SEVERAL POINTS:

(I) HONEYWELL IS NOT ENTITLED TO THE OPPORTUNITY TO RESUBMIT ON GROUNDS OF FAIRNESS;

(II) IN VIEW OF THE FULL OPPORTUNITY HONEYWELL AND THE OTHERS HAD TO SUBMIT ACCEPTABLE PROPOSALS IN THE FIRST INSTANCE, ONE OR MORE OF THEM, AND THE COMPLETE INFORMATION THAT ALL COMPETITORS HAD ABOUT THE REQUIREMENTS AND THE TYPE OF LIVE TEST TO BE RUN, THE AIR FORCE WAS JUSTIFIED IN DECIDING THAT IT WOULD HAVE SLIGHT ASSURANCE THAT ANY REASONABLE EXTENSION WOULD YIELD ADDITIONAL ACCEPTABLE PROPOSALS;

(III) THE 200 HOUR REQUIREMENT REPRESENTED A VALID AIR FORCE NEED;

(IV) THE ESTIMATED LIFE CYCLE COSTS OF THE IBM PROPOSAL COMPARE FAVORABLY WITH THOSE FOR THE PROTESTANT SO THAT THERE IS NOT THE SORT OF GROSS COST DISPARITY WHICH WOULD INDICATE IBM'S PROPOSAL WAS UNREASONABLY HIGH SO THAT EITHER THE 200 HOUR REQUIREMENT SHOULD HAVE BEEN RELAXED OR THE PROGRAM DELAYED IN THE HOPE THAT OTHER RESPONSIVE PROPOSALS MIGHT BE SUBMITTED;

(V) THE URGENT NEED TO MEET THE DATA PROCESSING REQUIREMENTS AT BASE LEVEL THROUGHOUT THE AIR FORCE;

(VI) ANY REOPENING OF THE COMPETITION WOULD HAVE INVOLVED UNACCEPTABLE DELAY.

THESE POINTS WILL BE DISCUSSED BELOW.

THE EFFORTS MADE BY THE AIR FORCE TO INSURE COMPLETE UNDERSTANDING OF AND RESPONSIVENESS TO THE REQUIREMENTS OF PHASE II WERE UNPRECEDENTED. ADDITION TO THE USUAL LETTER OF INTEREST ISSUED ON MAY 29, 1966 TO 44 VENDORS, TWO OTHER PRELIMINARY LETTERS WERE ADDRESSED TO COMPANIES ON THE MASTER BIDDERS LIST. THE FIRST OF THESE WAS MAILED ON JANUARY 21, 1966 BY ESQ. IT DISCUSSED THE STATUS OF PHASE II AND ESTABLISHED GUIDELINES TO BE FOLLOWED IN THESE AREAS: SYSTEMS, MODULARITY, OPERATING CONCEPTS, AND EQUIPMENT SELECTION AND IMPLEMENTATION. ON FEBRUARY 8, 1966, SECRETARY MARKS ADDRESSED A LETTER TO 32 COMPANIES IN WHICH PHASE II WAS DISCUSSED. IN THAT LETTER HE STRESSED HIS INTEREST IN THE "INTENT AND CAPABILITY OF THE COMPUTER INDUSTRY TO RESPOND TO OUR SPECIFIC NEEDS FOR -THIRD GENERATION- MODULARITY IN BOTH EQUIPMENT AND SOFTWARE WITHIN THE TIME CONSTRAINTS LAID DOWN. " IN MARCH 1966, ESQ MAILED ADDITIONAL ADVANCE INFORMATION CONCERNING AIR FORCE COBOL REQUIREMENTS. DURING 1966, SECRETARY MARKS AND THE AIR FORCE DIRECTOR OF DATA AUTOMATION VISITED THE LARGER MANUFACTURERS. IN THE COURSE OF THESE VISITS THE VENDORS WERE ENCOURAGED TO PRESENT THEIR VIEWS CONCERNING ANY ASPECT OF THE DATA AUTOMATION PROGRAM. AFTER ISSURANCE OF THE RFP THE COMPETITORS HAD ADEQUATE TIME TO SUBMIT THEIR PROPOSALS. THE RFP INVITED THEM TO SUBMIT ANY QUESTIONS THEY MIGHT HAVE ABOUT THE REQUIREMENTS OF THE RFP TO ESQ. ANSWERS TO SUCH INQUIRIES WERE PROVIDED TO ALL COMPETITORS. WHERE NECESSARY THE RFP WAS AMENDED. THE RFP CLEARLY SET FORTH THE METHOD OF TEST AND SELECTION, A METHOD WITH WHICH THE INDUSTRY WAS FULLY FAMILIAR. HONEYWELL WAS TREATED FAIRLY AND HAD EVERY OPPORTUNITY TO SUBMIT A RESPONSIVE PROPOSAL.

IBM'S PROPOSAL WAS THE ONLY ONE TO OFFER EQUIPMENT THAT AS TESTED PERFORMED IN ACCORDANCE WITH THE AIR FORCE MANDATORY REQUIREMENTS. UNDER THESE CIRCUMSTANCES CONSIDERATION OF COST DOES NOT ENTER INTO THE EVALUATION PROCESS PRESCRIBED BY BOB CIRCULAR NO. A-54 AND DOD DIRECTIVE 4105.55. COST IS RELEVANT ONLY, AS DISCUSSED BELOW, TO A DETERMINATION WHETHER THE COMPETITION SHOULD BE REOPENED TO THE ORIGINAL PROPOSERS AND OTHERS OR THE MANDATORY REQUIREMENT SHOULD BE RELAXED.

THE FAIRNESS AND FULL OPPORTUNITY AFFORDED ALL PROPOSERS BY THE RFP CASTS CONSIDERABLE DOUBT UPON ANY CLAIMS THAT PROPOSALS COULD HAVE BEEN SATISFACTORILY AMENDED WITHIN A SHORT PERIOD OF TIME. PAST EXPERIENCE WITH ADPE SELECTIONS HAD INDICATED THAT VENDOR EAGERNESS WILL PRODUCE CLAIMS OF PERFORMANCE THAT CANNOT BE MET. IT IS PRECISELY FOR THIS REASON THAT LIVE TEST DEMONSTRATIONS ARE REQUIRED. IT IS ALSO FOR THIS REASON THAT IT IS AIR FORCE POLICY TO DISCUSS MINOR DEFICIENT ASPECTS OF OTHERWISE ACCEPTABLE SYSTEMS WITH THE VENDOR DURING EVALUATION ONLY WHEN THE PROBLEM IS READILY IDENTIFIABLE TO A SPECIFIC HARDWARE OR SOFTWARE DEFICIENCY THAT CAN EASILY BE CURED. HONEYWELL'S DEFICIENCY WAS NOT OF THIS SORT. AS ANALYZED BY THE AIR FORCE ANY ATTEMPT TO REVISE THE HONEYWELL PROPOSAL WOULD REQUIRE A COMPLETE RESUBMISSION OF ALL TIMING TABLES AND A SECOND LIVE TEST DEMONSTRATION, A TIME CONSUMING PROCESS GIVING NO ASSURANCE THAT THE REVISED PROPOSAL COULD MEET THE 200 HOUR REQUIREMENT. IT WAS THE CONCLUSION OF AIR FORCE TECHNICAL EXPERTS THAT THE DEGRADATION OF PROCESSING TIMES CONTAINED IN THE PROPOSAL, BROUGHT TO LIGHT DURING THE LIVE TEST DEMONSTRATION, WAS CAUSED BY UNUSUAL SOFTWARE AND HARDWARE CONVENTIONS EMPLOYED BY HONEYWELL IN PROCESSING REAL TIME TRANSACTIONS. THESE SOFTWARE CONVENTIONS ARE DEEPLY EMBEDDED ELEMENTS OF A COMPLEX ASSEMBLY OF INTEGRATED SOFTWARE CONVENTIONS THAT MAKE UP WHAT IS GENERALLY REFERRED TO AS THE OPERATING/MONITOR/EXECUTIVE CONTROL SYSTEM. EXPERIENCE WITHIN THE AIR FORCE AND THE COMPUTER INDUSTRY GENERALLY INDICATES THAT SUCH SYSTEMS ARE EXTREMELY COMPLEX IN DESIGN AND UNUSUALLY TIME CONSUMING IN DEVELOPMENT.

THE CENTRAL PROCESSING UNIT OF THE HONEYWELL PROPOSED SYSTEM WHEN COMPARED TO THE EQUIPMENT OF THE OTHER THREE OFFERORS WHICH CAN WITHOUT QUESTION BE CATEGORIZED AS "THIRD GENERATION" IS SUBSTANTIALLY SLOWER. HONEYWELL'S PROCESSING SPEED IS FURTHER DEGRADED BY THE NECESSITY OF PROCESSING EIGHT BIT CHARACTER REPRESENTATIONS WITHIN A CENTRAL PROCESSING UNIT DESIGNED TO PROCESS SIX BIT CHARACTER REPRESENTATIONS. IT WAS THEN THE CONSIDERED OPINION OF THOSE EVALUATING HONEYWELL'S PROPOSAL THAT THE DEFICIENCIES IN IT WERE SUBSTANTIAL AND COMPLEX, NOT MATTERS EASILY REPAIRED. THE 200 HOUR MANDATORY REQUIREMENT WAS REEXAMINED WHEN IT WAS DETERMINED THAT ONLY ONE OFFEROR WAS RESPONSIVE. THE REEXAMINATION VERIFIED THAT THE REQUIREMENT WAS SOUND. THE CAPABILITY DEFINED BY THE 200 HOURS IS NEEDED TO HANDLE CURRENT WORKLOAD WITH STAFFING FOR TWO EIGHT HOUR SHIFTS PER DAY, FIVE DAYS A WEEK, AND ANTICIPATED WORKLOAD GROWTH THROUGHOUT THE SIX YEAR LIFE OF THE SYSTEM WITH ADDITIONAL STAFFING, BUT WITHOUT REQUIRING ADDITIONAL EQUIPMENT. TO RELAX THE REQUIREMENT WOULD MEAN THAT THE AIR FORCE WOULD ACQUIRE AND INSTALL EDPE THAT WOULD BE NEAR SATURATION AT THE TIME OF INSTALLATION.

A SUBSIDIARY ISSUE TO HONEYWELL'S CONTENTION THAT IBM SHOULD NOT HAVE BEEN SELECTED WITHOUT AN OPPORTUNITY BEING GIVEN TO THE NON RESPONSIVE OFFERORS TO REPAIR THEIR PROPOSALS, BUT PROBABLY THE MOST DRAMATIC ALLEGATION OF THE PROTEST, IS THE ASSERTION THAT THE SELECTION OF IBM WILL RESULT IN AN ADDITIONAL EXPENDITURE BY THE GOVERNMENT OF SOME $60 MILLION. THIS FIGURE, FOCUSED ON BY HONEYWELL AND OTHERS CRITICIZING THE SELECTION, IS BASED SOLELY ON A CONSIDERATION OF EQUIPMENT ACQUISITION COST AND FAILS TO MEASURE THE PERFORMANCE CAPABILITY OF THE SYSTEM--- ITS COST EFFECTIVENESS. IT SHOULD BE NOTED THAT BOB CIRCULAR NO. A-54 DIRECTS ATTENTION TO OVERALL COST TO THE GOVERNMENT. AFTER IBM'S PROPOSAL WAS ISOLATED AS THE ONLY RESPONSIVE ONE, THE AIR FORCE SOUGHT TO DETERMINE WHETHER THE OVERALL SYSTEM COST OF THAT PROPOSAL WAS REASONABLE. UNREASONABLE THE PROPER STEP WOULD NOT HAVE BEEN SELECTION OF HONEYWELL BUT CANCELLATION OF THE RFP AND RESOLICITATION. FROM THE POINT OF VIEW OF COST BENEFIT, THERE WAS NO DOUBT ABOUT THE PROPOSAL'S REASONABLENESS. "COST BENEFIT" STUDY CONDUCTED IN LATE 1965 HAD DETERMINED THAT A PHASE II SYSTEM COULD COST UP TO $250 MILLION AND STILL PROVIDE A NET COST BENEFIT VS. CURRENT SYSTEMS.

A COMPARISON IN TOTAL SYSTEMS COST FOR THE SIX YEAR LIFE WAS MADE OF THE IBM AND HONEYWELL PROPOSALS. IN MAKING THIS COMPARISON THE EFFECT OF ANTICIPATED WORKLOAD GROWTH ON EACH SYSTEM WAS CALCULATED. HONEYWELL'S CONFIGURATION WAS SATURATED SHORTLY AFTER INSTALLATION. OPERATION AND MAINTENANCE COSTS WHICH WOULD BE INCURRED WHEN THE PROCESSING REQUIREMENTS REACHED THE LEVEL WHERE SUCH COSTS WERE NO LONGER COVERED BY THE VENDORS' PROPOSALS WERE CALCULATED. WHEN A VENDOR'S PROPOSED CONFIGURATION WAS UNABLE TO PROCESS THE EXPANDED WORKLOAD WITHIN THE MAXIMUM MANNING PERIOD (24 HOURS PER DAY), AN EXPANSION OF THE CONFIGURATION WAS REQUIRED. THE COST OF SUCH FORESEEABLE EXPANSION WAS INCLUDED IN THE ESTIMATED TOTAL SYSTEM COSTS. THIS STUDY ESTABLISHED THAT THE LIFE CYCLE COST OF THE IBM PROPOSED SYSTEM WAS COMPARABLE TO THE COST OF THE HONEYWELL PROPOSED CONFIGURATION WITH THE ADDITIONAL INCREMENTS OF EQUIPMENT NEEDED TO HANDLE THE ANTICIPATED WORKLOAD THROUGHOUT THE SIX YEAR LIFE.

HONEYWELL ARGUES THAT THE FACT THAT SUCH A COST COMPARISON WAS CONDUCTED IS INCONSISTENT WITH THE DETERMINATION THAT ITS PROPOSAL WAS NON- RESPONSIVE. AS ALREADY STATED, THE STUDY WAS NOT CONDUCTED TO ASCERTAIN IF HONEYWELL'S PROPOSAL SHOULD BE ACCEPTED, BUT TO VALIDATE THE REASONABLENESS OF THE COST OF THE ONLY RESPONSIVE PROPOSAL, THAT OF IBM. ARGUMENT II

HONEYWELL CONTENDS THAT IN THE EVENT THAT IT IS DECIDED THAT NEGOTIATIONS WERE NOT NECESSARY, THEIR PROPOSAL SHOULD NONETHELESS BE CONSIDERED RESPONSIVE SINCE IT OFFERED TO DELIVER A SYSTEM WHICH WOULD OPERATE WITHIN THE 200 HOUR WORKLOAD REQUIREMENTS. AIR FORCE RESPONSE

IN PAST YEARS, IT HAS BEEN THE AIR FORCE EXPERIENCE THAT WRITTEN STATEMENTS ABOUT EQUIPMENT CAPABILITY FREQUENTLY WERE PROVEN INVALID WHEN THE EQUIPMENT WAS ACTUALLY INSTALLED. THIS EXPERIENCE IS NOT LIMITED TO THE AIR FORCE. SINCE WITH ADPE PROGRAMS NO FIRM QUANTITY CONTRACT WITH BINDING PERFORMANCE COMMITMENTS FOLLOWS SELECTION, THE PENALTIES SUFFERED BY AN OVEROPTIMISTIC MANUFACTURER ARE MINIMAL. DUE TO THE UNDERSTANDABLE TENDENCY OF OFFERORS TO UFF," THE AIR FORCE AND THE DEPARTMENT OF DEFENSE INSTITUTED A POLICY THAT A PRESELECTION DEMONSTRATION OR FULL SCALE TEST, OF EQUIPMENT CAPABILITIES WOULD BE A MANDATORY REQUIREMENT IN ALL EDP EQUIPMENT SELECTIONS. THIS POLICY IS NOW CONTAINED IN DOD DIRECTIVE 4105.55. THE LIVE TEST DEMONSTRATION OF HONEYWELL'S PROPOSED SYSTEM FAILED TO SUBSTANTIATE THE WRITTEN PROPOSAL'S DESCRIPTION OF ITS PERFORMANCE. THE VERY PURPOSE OF THE LIVE TEST DEMONSTRATION IS TO ASCERTAIN WHETHER THE PERFORMANCE DESCRIPTIONS IN THE PROPOSAL ARE ACTUALLY MET BY THE EQUIPMENT OFFERED. THE FAILURE OF HONEYWELL'S EQUIPMENT TO DEMONSTRATE REQUIRED SPEED AND CAPABILITY DURING THE LIVE TEST CONSTITUTED SUFFICIENT BASIS TO HOLD THAT ITS PROPOSAL FAILED TO MEET THE MANDATORY REQUIREMENT OF THE RFP. IT'S SIGNIFICANT TO NOTE THAT HONEYWELL HAS ACKNOWLEDGED THAT ITS PROPOSED SYSTEM FELL SHORT OF THE MANDATORY 200 HOURS OPERATIONAL USE CRITERION. ARGUMENT III

HONEYWELL CONTENDS THAT A MANDATORY REQUIREMENT OF THE RFP WAS MATERIALLY CHANGED IN A LATER ADDENDUM DESIGNED TO BRING IBM WITHIN THE REQUIREMENTS. THIS ALLEGED PREFERENTIAL TREATMENT, IN HONEYWELL'S VIEW, RENDERS THE SELECTION OF IBM ILLEGAL. AIR FORCE RESPONSE

AT THE TIME THE RFP WAS ISSUED, ONE MANDATORY REQUIREMENT STATED THAT "AN ERROR CHECK FEATURE MUST BE INCLUDED TO VERIFY THAT THE DATA RECEIVED BY THE COMPUTER AGREES WITH THE DATA SENT. AUTOMATIC RETRANSMISSION MUST BE PROVIDED AS A MEANS OF ERROR CORRECTION. (RFP, ATTACHMENT 2, PARA. 5-2-9H (5) ).' THIS REQUIREMENT WAS ELIMINATED AS A MANDATORY REQUIREMENT OF THE RFP BY ADDENDUM 25-65-A5 ISSUED 5 OCTOBER 1966. THIS ADDENDUM WAS NOT ISSUED TO AID IBM OR ANY OTHER VENDOR. IT WAS ISSUED UNDER THE FOLLOWING CIRCUMSTANCES. DURING THE DEVELOPMENT OF THE DATA SYSTEM SPECIFICATION CONSIDERABLE EMPHASIS WAS PLACED ON THE NEED FOR RELIABLE DATA VIA REMOTE INPUT/OUTPUT DEVICES. THE AIR FORCE WAS PRIMARILY CONCERNED WITH ERROR DETECTION AND CORRECTION HARDWARE AND SOFTWARE TECHNIQUES THAT WOULD INSURE A HIGH DEGREE OF VALIDITY IN THE TRANSMISSION OF DATA TO AND FROM THE COMPUTER VIA TELEPHONE LINES AND REMOTE INPUT/OUTPUT DEVICES. THERE WAS NO QUESTION OF THE NEED FOR ERROR DETECTION WHICH WAS DETERMINED A "MUST.' WITH RESPECT TO ERROR CORRECTION THERE WAS MUCH DISCUSSION WHICH FUNDAMENTALLY CENTERED ON TWO POINTS:

(I) HOW NECESSARY ARE ERROR-CORRECTION TECHNIQUES TO THE CURRENT STATE-OF -THE-ART HARDWARE CONFIGURATIONS AND ADVANCED COMMUNICATION FACILITIES?

(II) SHOULD THE AIR FORCE PAY THE POSSIBLE SUBSTANTIAL COSTS RESULTING FROM SUCH A EQUIREMENT? A REVIEW OF THE TECHNICAL INFORMATION AVAILABLE INDICATED THAT MOST MANUFACTURERS HAD THE CAPABILITY, OR COULD PROVIDE IT, WITHIN THE TIME CONSTRAINTS OF THE RFP. IT WAS DETERMINED THAT THE BEST INTEREST OF THE AIR FORCE WOULD BE SERVED TO STRIVE FOR MAXIMUM "ERROR- FREE" DATA TRANSMISSION AND, ACCORDINGLY, A MANDATORY REQUIREMENT WAS INCLUDED ON "ERROR CORRECTION.' THE AIR FORCE WAS ALSO CONCERNED ABOUT PROBLEMS WHICH WOULD ENSUE BY NOT SPECIFYING "ERROR-CORRECTION" AS A MANDATORY REQUIREMENT.

SHORTLY AFTER THE RFP HAD BEEN ISSUED, THE AIR FORCE RECEIVED SEVERAL INFORMAL QUESTIONS BY TELEPHONE FROM PARTICIPATING VENDORS REGARDING INTERPRETATION OF REMOTE INPUT/OUTPUT DEVICE SPECIFICATIONS CONTAINED IN PARAGRAPH 5-2-9H OF ATTACHMENT II OF THE RFP. MOST PARTIES SOUGHT CLARIFICATION AS TO THE DEFINITION OF THE TERM"ERROR-CORRECTION" AND OTHER TERMS CONTAINED IN THE REFERENCED PARAGRAPH. ULTIMATELY, IN ACCORDANCE WITH STANDARD PROCEDURES, THESE TELEPHONE INQUIRIES WERE SUBMITTED AS FORMAL WRITTEN QUESTIONS BY THE PARTICIPATING VENDORS. IN THE MEANTIME, FURTHER INVESTIGATION AND STUDY OF ERROR DETECTION AND CORRECTION TECHNIQUES WERE CARRIED OUT BY THE AIR FORCE INCLUDING CONSULTATION WITH GEEIA PERSONNEL LOCATED AT GRIFFISS AFB, NEW YORK. ALSO AN INFORMAL COST STUDY UTILIZING REPORTS PREPARED BY AUERBACH, INC., WAS ACCOMPLISHED TO DETERMINE HOW MUCH THE AIR FORCE WOULD BE PAYING FOR "ERROR-CORRECTION" TECHNIQUES.

IT WAS CONCLUDED THAT THE COST OF A REMOTE DEVICE WITH "ERROR CORRECTION" CAPABILITY WOULD BE ROUGHLY DOUBLE THE COST OF A REMOTE DEVICE WITHOUT THIS CAPABILITY. THIS INFORMATION, AS TO COST, TOGETHER WITH NEW INTELLIGENCE GATHERED FROM THE GEEIA AIR FORCE OFFICE AS TO NEW METHODS OF ERROR CHECKING BEING EMPLOYED BY THE COMPUTER INDUSTRY, CAUSED THE AIR FORCE TO REFLECT UPON AND RECONSIDER THE NECESSITY OF "ERROR-CORRECTION" AS A MANDATORY REQUIREMENT OF THE RFP. WE BELIEVE THIS TO BE A PRUDENT AND JUST DECISION. TO HAVE PAID AN ADDITIONAL SEVERAL MILLION DOLLARS FOR A NICE-TO-HAVE FEATURE COULD WELL HAVE REFLECTED UNFAVORABLY ON THE AIR FORCE. ACCORDINGLY, THE REQUIREMENT FOR "ERROR-CORRECTION" WAS ELIMINATED AS A MANDATORY REQUIREMENT OF THE RFP BY ESQ ADDENDUM 25-65-A5 ISSUED 5 OCTOBER 1966.

UPON DISSEMINATION OF THE REFERENCED ADDENDUM, HONEYWELL OBJECTED SOMEWHAT STRONGLY ON THE GROUND THAT THE CHANGE IN REQUIREMENTS CONSTITUTED A WAIVER OF THIS MANDATORY REQUIREMENT ALLEGING THAT THE CHANGE SERIOUSLY AFFECTED ITS MARKETING STRATEGY. HONEYWELL WAS ADVISED THAT THE AIR FORCE WAS NOT THEN, OR PREVIOUSLY, AWARE OF THE COMPANY'S MARKETING STRATEGY, OR THAT OF ANY OTHER PARTICIPATING VENDOR, AS IT APPLIED TO REMOTE INPUT/OUTPUT DEVICES TO BE PROPOSED.

IT WAS FURTHER EXPLAINED THAT THE DECISION REACHED WAS A RECONSIDERATION BY THE AIR FORCE BASED ON REDUCING THE OVERALL COST OF THE PROGRAM. WAS EMPHASIZED THAT "ERROR-CORRECTION" TECHNIQUES WERE STILL CONSIDERED A MOST DESIRABLE EQUIPMENT ATTRIBUTE BY THE AIR FORCE, AND SHOULD BE PROPOSED IF GOOD COST/PERFORMANCE TRADE-OFFS WERE POSSIBLE.

HONEYWELL WAS FURTHER INFORMED THAT, SHOULD THERE BE A QUESTION OF GOOD COST/PERFORMANCE TRADE-OFF, THE AIR FORCE ENCOURAGED ALTERNATE PROPOSALS. THE SUGGESTION BEING MADE WAS THAT HONEYWELL SHOULD SUBMIT TWO PROPOSALS: ONE WITH "ERROR-CORRECTION" AND ONE WITHOUT. SUCH ALTERNATE SUBMISSIONS ARE NOT EXPENSIVE TO MAKE. WITH THEM THE AIR FORCE WOULD BE IN A POSITION TO JUDGE THE DEGREE OF COST/PERFORMANCE TRADE-OFF POSSIBLE WITH BOTH TYPES OF EQUIPMENT. HONEYWELL REPLIED THAT THEY WOULD CONSIDER SUCH A MARKETING POSITION. HOWEVER, IN THE IMMEDIATE TIME PERIOD THAT FOLLOWED, HONEYWELL APPEALED TO THE AIR FORCE TO REVERSE THE DECISION TO ELIMINATE THIS MANDATORY REQUIREMENT.

THE AIR FORCE DECISION TO ELIMINATE THIS MANDATORY REQUIREMENT WAS NOT ARRIVED AT HASTILY; DUE REGARD AND CONSIDERATION WERE ACCORDED TO THE PARTICIPATING VENDORS AND THE INTEREST OF THE AIR FORCE. IT WAS CONSIDERED THAT THE REMOVAL OF THIS MANDATORY REQUIREMENT WOULD NOT UNDULY AFFECT ANY ONE OF THE PARTICIPATING VENDORS OR ANY VENDOR WHO HAD CHOSEN NOT TO PARTICIPATE. AS OF 4 OCTOBER 1966, SUFFICIENT TIME REMAINED PRIOR TO THE SUBMISSION DATE OF PROPOSALS TO PERMIT THIS CHANGE TO REMOTE DEVICES. THIS CHANGE IN THE MANDATORY REQUIREMENTS DID NOT PREVENT A PARTICIPATING VENDOR FROM PROPOSING "ERROR CORRECTION" TECHNIQUES, AS TWO VENDORS DID. HONEYWELL DID NOT, IN THE END, PROPOSE "ERROR-CORRECTION" TECHNIQUES AS A CONSIDERED PART OF THE PROPOSAL SUBMITTED, NOR DID THIS VENDOR SUBMIT AN ALTERNATE PROPOSAL CONTAINING SUCH TECHNIQUES. THIS LEAVES ONE TO CONCLUDE THAT THE CHANGE IN MANDATORY REQUIREMENT DISCUSSED ABOVE DID NOT HAVE A SERIOUS, ADVERSE EFFECT ON HONEYWELL'S PROPOSAL.

THE ALLEGATION MADE IN THE PROTEST THAT THE DELETION OF THIS MANDATORY REQUIREMENT WAS ,DESIGNED TO BRING IBM WITHIN THE REQUIREMENTS IN A MANNER MOST ADVANTAGEOUS TO IBM" IS COMPLETELY WITHOUT FOUNDATION. THE AIR FORCE CONSIDERS THE CHANGE FROM "MANDATORY" TO "DESIRABLE" AS COMPLETELY JUSTIFIED IN THE LIGHT OF COST CONSIDERATIONS. ARGUMENT IV

HONEYWELL CONTENDS THAT IF FUTURE NEGOTIATIONS WITH HONEYWELL ARE NOT CONDUCTED, THE PROCUREMENT SHOULD BE CANCELLED AND NEW ACTION SHOULD BE INITIATED TO OBTAIN A CONTRACT FOR A REASONABLE PRICE. AIR FORCE RESPONSE

THIS POINT IS ADDRESSED FULLY IN THE RESPONSES TO ARGUMENT I AND HONEYWELL'S REQUESTED REOPENING OF COMPETITION. ARGUMENT V

A FINAL ARGUMENT, POSED IN QUESTION FORM BY HONEYWELL, CONCERNS IBM'S RESPONSIVENESS TO THE RFP. HONEYWELL BELIEVES IT IS NOT POSSIBLE, ASSUMING THAT IBM PROPOSED "SOME FORM OF MULTI-PROCESSING AT THE -D- LEVEL," TO RECONCILE ITS EQUIPMENT WITH THE MANDATORY REQUIREMENTS FOR MAGNETIC TAPE UNITS. AIR FORCE RESPONSE

HONEYWELL'S QUESTION IS WHETHER THE WINNING OFFEROR PROPOSED A MULTI- PROGRAMMING MODE OF PROCESSING IN MEETING THE REQUIREMENTS OF THE RFP. THE TERM MULTI-PROCESSING IS USED IN THE PROTEST; HOWEVER, THE AIR FORCE USES THE TERM MULTI-PROGRAMMING TO DESCRIBE THE MODE OF OPERATION THAT HAS BEEN ADDRESSED.

THE ANSWER TO HONEYWELL'S QUESTION IS "YES.' BUT ITS INFERENCE THAT IBM COULD NOT BE RESPONSIVE IF THE PROPOSED EQUIPMENT IS LIMITED TO FIVE MAGNETIC TAPE UNITS AND AT THE SAME TIME OFFERS AMULTI PROGRAMMING MODE OF OPERATION IS FALSE. THE WINNING VENDOR DID IN FACT PROPOSE FIVE MAGNETIC TAPE UNITS AND MULTI-PROGRAMMING. THE

THE HONEYWELL ASSUMPTION THAT IBM MUST BE NONRESPONSIVE IF ITS AIR FORCE HAS DETERMINED THAT THE IBM PROPOSAL WAS RESPONSIVE. PROPOSAL COMBINES FIVE TAPE UNITS AND MULTI-PROGRAMMING STEMS FROM THEIR LACK OF COMPLETE KNOWLEDGE ABOUT THAT PROPOSAL. AS STATED IN THE COVER LETTER THIS INFORMATION ABOUT THE IBM PROPOSAL IS PROPRIETARY AND RELEASE TO HONEYWELL WOULD GIVE THEM AN UNFAIR COMPETITIVE ADVANTAGE. HONEYWELL'S REQUESTED REOPENING OF COMPETITION HONEYWELL HAS REQUESTED THAT YOU DECLARE THE SELECTION OF IBM TO BE TENTATIVE AND THAT THEY (AND PRESUMABLY OTHER PROPOSERS) BE GIVEN THE OPPORTUNITY TO SUBMIT REVISED PROPOSALS. AIR FORCE RESPONSE

FOR THE REASONS SET FORTH IN RESPONSE TO THE INDIVIDUAL ARGUMENTS OF HONEYWELL, THE AIR FORCE DOES NOT BELIEVE THAT FAIRNESS REQUIRES OR THAT IT WOULD BE IN THE BEST INTEREST OF THE GOVERNMENT TO REOPEN THIS COMPETITION.

THE BASE LEVEL DATA AUTOMATION PROGRAM IS NOW IN PROCESS OF IMPLEMENTATION. THE DATA SYSTEM DESIGN CENTER WITHIN THE AIR FORCE IS PROCEEDING WITH THE DETAILED SYSTEMS DESIGN AND PROGRAMMING OF DATA SYSTEMS UTILIZING THE DISCRETE SOFTWARE LANGUAGES AND CONVENTIONS OF THE WINNING PROPOSER. BOTH THE AIR FORCE AND THE WINNING MANUFACTURER ARE PROCEEDING WITH THE DEVELOPMENT OF PLANS FOR DELIVERY AND INSTALLATION OF EQUIPMENT. IBM PLANS TO DELIVER AND INSTALL TWO SYSTEMS IN JULY 1967 AT ITS EXPENSE. THESE SYSTEMS ARE TO BE UTILIZED FOR TRAINING AND THE PROGRAM TEST AND DEBUGGING OF COMPUTER PROGRAMS NOW BEING DEVELOPED. SHORT, IT CAN BE STATED THAT THE AIR FORCE IS PROCEEDING TO IMPLEMENT THIS PROGRAM. SERIOUS AND COSTLY DELAYS WOULD ENSUE IF THE AIR FORCE WERE TO PURSUE OTHER THAN THIS PRESENT COURSE OF ACTION.

IT IS IMPERATIVE THAT AIR FORCE BASE LEVEL MANAGERS BE PROVIDED WITH MODERN AND RESPONSIVE EQUIPMENT NECESSARY TO THE EFFICIENT MANAGEMENT OF AIR FORCE MISSIONS. FURTHER, THE COST BENEFITS ACHIEVABLE THROUGH THE INSTALLATION OF THE WINNING OFFEROR'S EQUIPMENTS MUST BE REALIZED AT THE EARLIEST POINT IN TIME. THESE COST BENEFITS IN THEMSELVES WILL OFFSET THE COST OF THE EQUIPMENT PROPOSED BY THE WINNING VENDOR.

IF, NOTWITHSTANDING THE CONDITIONS STATED ABOVE, THE AIR FORCE WERE TO RECONSIDER THE SELECTION OF THE WINNING PROPOSAL AND REISSUE THE REQUEST FOR PROPOSAL, THE OVERALL SLIPPAGE IN THIS PROGRAM WOULD UNDOUBTEDLY EXCEED TWELVE MONTHS IN TIME WITH A RESULTING LOSS OF TREMENDOUS COST BENEFITS.

HONEYWELL'S ATTORNEYS WERE GIVEN AN OPPORTUNITY TO EXAMINE AND REPLY TO THE REPORT ON AIR FORCE PHASE II SELECTION. HONEYWELL'S REPLY IS A RATHER LENGTHY LETTER, WITH NINE ENCLOSURES, DATED JUNE 28, 1967, IN WHICH ISSUE IS TAKEN WITH THE AIR FORCE POSITION ON MOST OF THE ABOVE POINTS. HOWEVER, NO REBUTTAL HAS BEEN MADE TO THE AIR FORCE POSITION ON THE ALLEGATIONS CONCERNING THE REQUIREMENT FOR AUTOMATIC RETRANSMISSION AND IBM'S RESPONSIVENESS TO THE MULTI-PROGRAMMING MODE OF OPERATION AT THE D LEVEL. WE BELIEVE THE AIR FORCE REPLY TO THESE ALLEGATIONS SATISFACTORILY ANSWERS THEM.

HONEYWELL MAKES A STRONG DEFENSE OF THE OTHER GROUNDS OF ITS PROTEST. HONEYWELL TAKES THE POSITION THAT ASSUMING, ARGUENDO, THE CORRECTNESS OF THE BENCHMARK TIMINGS OF SOME 260 HOURS FOR THE C AND D LEVELS, THE AIR FORCE IS STILL REQUIRED TO ENGAGE IN DISCUSSIONS WITH IT TO DETERMINE AT WHAT ADDITIONAL COST AND ON THE BASIS OF WHAT TIME SCHEDULE IT CAN MEET THE TIME REQUIREMENT. THIS IS SO, HONEYWELL SAYS, BECAUSE THE REQUIREMENT OF THE STATUTE APPLIES TO ALL PROPOSALS WITHIN "COMPETITIVE RANGE, PRICE, AND OTHER FACTORS CONSIDERED.' THE STATUTE THEREFORE REQUIRES A CONSIDERATION OF BOTH PRICE AND TECHNICAL COMPLIANCE IN DETERMINING "COMPETITIVE RANGE.' THUS INTERPRETED, THE ASSERTED TECHNICAL DEFICIENCIES ALONE DO NOT JUSTIFY THE FAILURE TO NEGOTIATE IN VIEW OF THE POSSIBLE SAVINGS OF SOME $60,000,000.

HONEYWELL ALSO TAKES ISSUE WITH THE AIR FORCE ARGUMENT THAT IT IS NOT NEGOTIATING A CONTRACT, BUT MERELY SELECTING A SOURCE FOR THE EQUIPMENT PURSUANT TO PUBLIC LAW 89-306, 40 U.S.C. 759, AND THAT ANY CONTRACT NEGOTIATED WILL BE BY GSA UNDER THE FEDERAL SUPPLY SCHEDULE (FSS) SYSTEM. HONEYWELL CONTENDS THAT ALL TERMS AND CONDITIONS OF ANY CONTRACT ARE NEGOTIATED BY THE AIR FORCE IN SELECTING THE SOURCE AND THESE TERMS AND CONDITIONS BECOME PART OF THE SELECTED SOURCE'S FSS CONTRACT BY MEANS OF A PRO FORMA PROCEDURE. PARAGRAPHS 2-4 AND 1-15 OF ATTACHMENT I OF THE RFP ARE CITED IN THIS CONNECTION. HONEYWELL ALSO POINTS TO THE FACT THAT A PRICE GUARANTEE OF MORE THAN 2 YEARS IS REQUIRED UNDER THE SUBJECT RFP, WHEREAS FSS CONTRACTS ARE FOR A YEAR, AS INDICATING THE AIR FORCE IS IN ACTUALITY DOING THE NEGOTIATING. ANOTHER REASON CITED IN SUPPORT OF ITS POSITION THAT 10 U.S.C. 2304(G) IS APPLICABLE TO THE SELECTION PROCEDURE IN THE INSTANT CASE IS THE FACT THAT ALTHOUGH THE STATUTE STATES CERTAIN EXCEPTIONS, NONE OF THEM ARE APPLICABLE HERE. HONEYWELL ALSO TAKES ISSUE WITH SEVERAL STATEMENTS CONCERNING THE NATURE AND EFFECT OF THE FSS SYSTEM, WHICH WE DO NOT BELIEVE ARE PARTICULARLY RELEVANT TO THE ISSUE AND THEREFORE, ARE NOT RESTATED HEREIN. HONEYWELL ALSO CITES OUR DECISION OF AUGUST 8, 1962, B-148889, STATING THAT IT HOLDS THAT THE ARMED SERVICES PROCUREMENT ACT APPLIES TO CONTRACTS MADE BY THE DEPARTMENT OF DEFENSE UNDER THE FSS SYSTEM.

HONEYWELL GOES TO SOME LENGTH IN AN EFFORT TO REFUTE THE AIR FORCE POSITION ON THE NATURE, EXTENT AND COMPLEXITY OF THE DEFICIENCIES WHICH CAUSED ITS SYSTEM TO FAIL THE BENCHMARK. AT THE SAME TIME, IT REITERATES ITS ARGUMENT THAT EVEN ASSUMING THE AIR FORCE IS CORRECT IN ITS TECHNICAL FINDINGS, DISCUSSIONS ARE REQUIRED UNDER 10 U.S.C. 2304 (G). BASICALLY, HONEYWELL ARGUES THAT THE AIR FORCE DOES NOT KNOW THE CAUSES OF THE DEGRADATION OR THE EXTENT AND NATURE OF THE CORRECTIONS NECESSARY TO BRING THE TIMINGS WITHIN THE 200 HOUR REQUIREMENT. FURTHERMORE, HONEYWELL ALLEGES THAT THE AIR FORCE HAS ARBITRARILY "SWITCHED" ITS POSITION ON THE NATURE OF THE DEFICIENCIES. TO SUPPORT THIS ARGUMENT, HONEYWELL CITIES CERTAIN AIR FORCE COMMENTS AT THE TIME OF THE DEBRIEFING AS INDICATING UNCERTAINTY AS TO THE NATURE OF THE DEFICIENCIES AND THE MUCH DIFFERENT POSITION NOW TAKEN TO DEFEND ITS FAILURE TO NEGOTIATE. THE FOLLOWING EXCHANGE OF CONVERSATION FROM THE DEBRIEFING IS QUOTED BY HONEYWELL AS AN EXAMPLE:

SPANGLE (HONEYWELL VICE PRESIDENT): COULD THE DEFICIENCY HAVE BEEN CORRECTED BY SOME ADDITIONAL EQUIPMENT OR WAS IT IN THE NATURE OF THE JOB AND OUR SYSTEM THAT THE ONLY CORRECTION WOULD HAVE BEEN MORE HOURS?

MURPHY (CIVILIAN ASSISTANT TO COLONEL STEFFES; MEMBER OF SOURCE SELECTION EVALUATION BOARD): THIS IS A VERY DIFFICULT CONCLUSION TO COME TO. WE EXPLORED THE POSSIBILITY AND WE COULDN-T COME TO A CONCRETE RECOMMENDATION AND WE WERE NOT ABLE TO MAKE ONE--- NEITHER THE EVALUATION OR THE ADVISORY COUNCIL.

SPANGLE: SO THE EXTRAPOLATION THEN TOOK THE FORM OF ADDING HOURS TO OUR PROPOSAL AND COSTS WERE DERIVED THEREFROM?

MURPHY: I WOULD SAY THAT THIS WAS THE PROPER CONCLUSION, YES.

SPANGLE: OF COURSE, IT MIGHT HAVE BEEN THAT HAD WE ADDED EQUIPMENT OUR COSTS STILL WOULD HAVE BEEN MUCH LOWER THAN THE SUCCESSFUL BIDDER.

MURPHY: AGAIN, I WOULD HAVE TO SAY IT MIGHT BE POSSIBLE--- YES.

STEFFES: IT MIGHT BE POSSIBLE THAT YOUR SOFTWARE MAY HAD TO HAVE BEEN MODIFIED TO A CERTAIN DEGREE. (DEBRIEFING Q-AND-A, ATTACH. 12, PP. 12- 13). AS A FURTHER EXAMPLE OF THE DIFFERENCE IN THE CONTEMPORANEOUS AND PRESENT POSITION, HONEYWELL CITIES THE FOLLOWING FROM THE DEBRIEFING:

IN ANALYZING THESE ADJUSTED TIMINGS FOR SYSTEM OPERATIONAL USE, THREE FUNCTIONAL PROCESSING CATEGORIES WERE IDENTIFIED; REAL-TIME, CONCURRENT BATCH, AND NON-CONCURRENT BATCH. THE ANALYSIS OF NON CONCURRENT BATCH PROCESSING, UNILATERALLY, WOULD INDICATE IT AS ACCEPTABLE.

THE DEGREE OF DEGRADATION TO CONCURRENT BATCH JOBS WHEN RUN, TOGETHER WITH REAL TIME PROCESSING, WAS DEMONSTRATED TO BE HIGHER THAN CONTAINED IN YOUR PROPOSAL. (THIS STATEMENT WAS REPEATED FOR EMPHASIS.) THE DEMONSTRATION OF THE REAL-TIME FUNCTION PROVIDED US WITH TIMINGS, WHICH WHEN ANALYZED, SHOWED A CONSIDERABLE INCREASE IN PROCESSING TIME OVER THAT WHICH WAS REPORTED. ALTHOUGH THE LIVE TEST DEMONSTRATION UNCOVERED DEFICIENCIES IN INPUT/OUTPUT ROUTINE HANDLING WITHIN THE OPERATING SYSTEM WHICH HAD A DEFINITE EFFECT ON RESPONSE TIMES TO REAL-TIME ACTIVITIES, A COMPLETE ANALYSIS WAS PERFORMED TO DETERMINE IF THERE WERE OTHER TECHNICAL REASONS CONTRIBUTING TO THE VARIATION IN THE REAL-TIME PROCESSING PER SE. OUR ANALYSIS SHOWED WHAT APPEARED TO BE A TIME-CONSUMING TECHNIQUE EMPLOYED IN THE AREA OF REMOTE MESSAGE HANDLING. ACCORDING TO YOUR TECHNICAL LITERATURE, WHICH WAS CONSIDERED EXCELLENT IN CLARITY AND COMPLETENESS, THE MASS STORAGE DEVICE IS UTILIZED IN LIEU OF CORE STORAGE FOR THE HANDLING OF INPUT/OUTPUT REMOTE MESSAGES WHICH REQUIRED ADDITIONAL CONTROL PROCESSING TO AND FROM THE MASS STORAGE DEVICE.

THIS TECHNIQUE TOGETHER WITH THE OPERATING SYSTEM DEFICIENCIES NOTED IN YOUR LETTER OF 10 FEBRUARY 1967 SUBSTANTIATED IN PART THE TIMINGS ACQUIRED AT THE LIVE TEST DEMONSTRATION FOR THE REAL-TIME PROCESSING. (ATTACHMENT 3, PP. 2-3) (ITALIC SUPPLIED.)

HONEYWELL DOES CONCEDE THAT THERE ARE HARDWARE AND SOFTWARE DEFICIENCIES, BUT CONTENDS THAT THEY ARE NOT COMPLEX AND INHERENT DEFICIENCIES, AS ALLEGED BY THE AIR FORCE, AND CAN BE READILY CORRECTED AT A COST OF APPROXIMATELY $1,250,000. THE DEFICIENCIES THEY IDENTIFY ARE IN USING MASS MEMORY INSTEAD OF CORE MEMORY AND AN OMISSION IN DRAWING UP THE SOFTWARE PACKAGE. THE TIMING DEGRADATION WAS PRIMARILY DUE TO THE OMISSION OF CERTAIN ACCESSES TO THE DISC FILE FOR EACH TRANSACTION IN DRAWING UP THE SOFTWARE PACKAGE, ACCORDING TO HONEYWELL.

HONEYWELL ALSO CHALLENGES THE AIR FORCE STATEMENTS THAT ITS EQUIPMENT IS NOT "THIRD NERATION," IS "SUBSTANTIALLY SLOWER," AND THE PROCESSING SPEED IS DEGRADED BY THE "NECESSITY OF PROCESSING EIGHT BIT CHARACTER REPRESENTATIONS WITHIN A CENTRAL PROCESSING UNIT DESIGNED TO PROCESS SIX BIT CHARACTER REPRESENTATIONS.' HONEYWELL HAS STATED ITS POSITION ON THESE MATTERS AS FOLLOWS:

(4) THE SUGGESTION THAT HONEYWELL'S PROPOSED SYSTEM IS NOT "THIRD GENERATION" IS CONTRARY TO THE POSITION OF RECOGNIZED AUTHORITIES IN THE DATA PROCESSING FIELD. IT IS PLAINLY WRONG. ATTACHMENTS 13 AND 14, STUDIES BY INDEPENDENT AUTHORITIES, PROVE THAT THE PROPOSED HONEYWELL SYSTEM IS "THIRD GENERATION.' THE AIR FORCE HAS RELIED UPON ONE OF THESE INDEPENDENT AGENCIES--- AUERBACH, INC.--- IN ITS DELETION OF THE AUTOMATIC RETRANSMISSION AS A MANDATORY REQUIREMENT (AF REP. 10).

(5) HONEYWELL'S EQUIPMENT IS NOT "SUBSTANTIALLY SLOWER.' THE FOLLOWING COMPARISON OF CENTRAL PROCESSOR SPEEDS ARE TAKEN FROM AUERBACK STANDARD EDP REPORTS, PAGES 11.210.110 AND 11.210.112, DATED JUNE 1966 (ATTACHMENT 15):

NECESSARY CYCLE TIME* ADD TIME

----------------- ----------------- HONEYWELL IBM LEVEL HONEYWELL IBM HONEYWELL IBM

1200 360/30 A 1.5 US 1.5 US 63 US 96 US

1200 360/30 B 1.5 US 1.5 US 63 US 96 US

1200 360/30 C 1.5 US 1.5 US 63 US 96 US

2200 360/40 D 1.0 US 1.25 US 51 US 64 US

*FOR 1 CHARACTER OR 1 BYTE (OR EQUIVALENT) OF INFORMATION. MOREOVER, WHEN MIXTURES OF TYPES OF OPERATIONS, REPRESENTATIVE OF BUSINESS DATA PROCESSING APPLICATIONS (SUCH AS ADD, COMPARE, MOVE AND BRANCH) ARE ANALYZED, THE RELATIVE SPEED STANDINGS OF IBM AND HONEYWELL ARE UNCHANGED. THUS THIS ASSERTION BY THE AIR FORCE IS CONCLUSIVELY REBUTTED BY INDEPENDENT AUTHORITY.

(6) THE AIR FORCE'S FURTHER CONTENTION THAT HONEYWELL'S PROCESSING SPEED IS DEGRADED BY THE ,NECESSITY OF PROCESSING EIGHT BIT CHARACTER REPRESENTATIONS WITHIN A CENTRAL PROCESSING UNIT DESIGNED TO PROCESS SIX BIT CHARACTER REPRESENTATIONS" IS ALSO FALSE AND REVEALS A BASIC FAILURE OF THE AIR FORCE TO UNDERSTAND THE HONEYWELL SYSTEM. EIGHT BIT CHARACTER REPRESENTATIONS ARE NOT PROCESSED. ON THOSE OCCASIONS WHEN EIGHT BIT CHARACTERS ARE INTRODUCED INTO THE SYSTEM, THEY PASS THROUGH A FAST ONE- TIME TRANSLATION; SINCE ALL INCOMING COMMUNICATIONS DATA IN EIGHT BIT FORM HAVE ONLY SIX SIGNIFICANT DATA BITS, ONLY SIX BIT DATA ARE ACTUALLY PROCESSED. THE STANDARD AIR FORCE CHARACTER SET CAN BE REPRESENTED IN SIX BITS. SINCE ALL SERIES 200 INTERNAL CODES ARE SIX BIT, FURTHER TRANSLATIONS IS UNNECESSARY. WHEN EIGHT BIT DATA IS REQUIRED ON OUTPUT, THE SAME FAST SIX TO EIGHT BIT TRANSLATION IS MADE. THE AIR FORCE IS CORRECT IN STATING THAT A DEGRADATION OF TIME OCCURS AS A RESULT. BUT THIS DEGRADATION IS MINUSCULE. USING AIR FORCE PARAMETERS FOR SIZES AND VOLUMES OF REAL-TIME, A DAILY ADDITION TO THE CENTRAL PROCESSOR TIME OF LESS THAN ONE-HALF MINUTE WOULD RESULT. FURTHERMORE, WERE TIME AVAILABLE WITHIN OPERATIONS, SUCH AS DISC ACCESSES, THIS TIME WOULD NOT BE ADDED TO THE TOTAL RUN TIME OF A JOB. COLLATERALLY, THE IBM 360 SYSTEM SELECTED CONFRONTS A SIMILAR DEGRADATION BECAUSE IT EMPLOYS VARIATIONS OF THE EIGHT -BIT BYTE IN MAKING UP ITS INTERNAL MACHINE CODES, REQUIRING A TRANSLATION OF 8 TO 6 OR 4 BITS.

AS TO THE "COST EFFECTIVENESS STUDY," HONEYWELL CHALLENGES IT ASA "GROSS RATIONALIZATION, FABRICATED UPON SPECULATION AND UNTENABLE ASSUMPTIONS.' HONEYWELL SAYS THAT IT IS IMPROPERLY BASED UPON THE ASSUMPTION THAT NO CORRECTION OF THE TIMING DEGRADATION IS POSSIBLE, AND THE FURTHER SPECULATIVE ASSUMPTION THAT IBM WILL NOT RAISE ITS PRICES AFTER THE 2-YEAR PRICE GUARANTEE. FURTHER, IT IS PREMISED ON THE MARGIN BY WHICH IBM IMPROVED UPON THE 200 HOUR REQUIREMENT, WHICH WAS NOT STATED AS A CONSIDERATION IN THE RFP. MOREOVER, IT IS ALLEGED THAT BOTH YOU AND ASSISTANT SECRETARY MARKS HAVE STATED SOME RESERVATIONS ABOUT THE VALIDITY OF THE STUDY. FINALLY, HONEYWELL QUESTIONS THE CONCLUSION THAT THE STUDY SHOWS THE COSTS OF THE TWO SYSTEMS ARE COMPARABLE SINCE THE FIGURES SHOW A DIFFERENCE OF SEVERAL MILLIONS OF DOLLARS. HONEYWELL ARGUES THAT THE AIR FORCE POSITION THAT DISCUSSIONS WITH IT WOULD AMOUNT TO ,UNFAIRNESS" TO IBM IGNORES THE FACT THAT NEGOTIATIONS PURSUANT TO 10 U.S.C. 2304 (G) FOR THE PURPOSE OF OBTAINING THE BEST POSSIBLE CONTRACT FOR THE GOVERNMENT IS EXACTLY WHAT THE STATUTE CONTEMPLATES AND REQUIRES. HONEYWELL ALSO POINTS OUT THAT ALL OFFERORS ARE ON NOTICE OF THE POSSIBILITY OF NEGOTIATIONS AND MODIFICATIONS OF PROPOSALS BY REASON OF THE STATUTE, PARAGRAPH 1-16-1, ATTACHMENT I, WHICH WAS QUOTED HERETOFORE, AND AIR FORCE PRACTICE IN PREVIOUS PROCUREMENTS.

FINALLY, HONEYWELL DISPUTES THE ASSERTION OF URGENCY ON THE PREMISE THAT THE EXISTING SYSTEMS ARE PRESENTLY PERFORMING THE NECESSARY FUNCTIONS, AND ACTUAL INSTALLATION OF THE OPERATIVE SYSTEM IS NOT SCHEDULED UNTIL MAY 1968.

BY LETTER DATED JULY 7, 1967, THE AIR FORCE HAS REPLIED TO HONEYWELL'S LETTER OF JUNE 28, AND HAS MADE SEVERAL STATEMENTS IN REBUTTAL. FIRST, IT IS POINTED OUT THAT BECAUSE OF THE GREATER CAPACITY OF THE IBM PROPOSAL, THAT IS, 160 HOURS TO HONEYWELL'S 260 HOURS, THE APPARENT INITIAL INVESTMENT COST DISPARITY OF SOME $60 MILLION DOES NOT STAND UP. THIS IS SO BECAUSE THE IBM PROPOSAL WILL PERMIT SATELLITING OF THE SMALLER BASES ON THE LARGER BASES, AS CONTEMPLATED IN THE RFP, AND FEWER THAN 135 SYSTEMS NEED TO INITIALLY PURCHASED. IN ADDITION, THIS PRINCIPLE WILL SERVE TO LOWER THE OVERALL COST FOR THE ANTICIPATED 6-YEAR LIFE.

WITH REGARD TO QUESTIONS HONEYWELL HAS RAISED ABOUT THE METHOD OF RELATING THE BENCHMARK DATA TO THE ACTUAL WORKLOAD, THE AIR FORCE REPORTS THAT THE BENCHMARK DEMONSTRATIONS WERE REPRESENTATIVE OF THE ACTUAL WORKLOADS AND EXTRAPOLATION WAS NECESSARY IN LIEU OF RUNNING THE ACTUAL MONTH-LONG WORKLOAD. FURTHERMORE, THE SAME TRANSACTIONS AND THE SAME METHOD OF EXTRAPOLATION WERE APPLIED TO ALL SYSTEMS AND THERE WAS A TREMENDOUS DIFFERENCE BETWEEN THE HONEYWELL AND IBM MONTHLY OPERATIONAL USE HOURS.

CONCERNING HONEYWELL'S ARGUMENT THAT FURTHER "DISCUSSIONS" WITH IT ARE REQUIRED BY 10 U.S.C. 2304 (G), THE AIR FORCE CONTENDS THAT EVEN IF THE STATUTE IS APPLICABLE TO THE SOURCE SELECTION PROCESS INVOLVED HERE, HONEYWELL IS NOT MERELY REQUESTING "DISCUSSION" BUT IS ASKING, IN EFFECT, THAT IT BE PERMITTED TO PROPOSE A DIFFERENT SYSTEM. THIS IS BEYOND THE LIMITS OF "COMPETITIVE RANGE," WHICH IS A MATTER TO BE DETERMINED BY THE AGENCY CONCERNED. IN THE INSTANT CASE, THE FACT THAT THE LIVE TEST DEMONSTRATION SHOWED THAT THE HONEYWELL SYSTEM FAILED BY A "SUBSTANTIAL ARGIN" JUSTIFIED THE CONCLUSION THAT IT WAS NOT WITHIN A COMPETITIVE RANGE. IF HONEYWELL IS PERMITTED TO REVISE ITS PROPOSAL TO THE EXTENT INDICATED, THE ARGUMENT IS MADE THAT OTHER OFFERORS WHOSE SYSTEMS DID NOT PERFORM AT ALL IN CERTAIN RESPECTS WOULD BE ENTITLED TO REVISE THEIR PROPOSALS, NECESSITATING VALIDATION AND LIVE TESTING AGAIN. THIS WOULD IN EFFECT UNDERCUT THE EDPE SELECTION PROCESS THAT HAS BEEN SUCCESSFULLY FOLLOWED IN SOME 30 PRIOR SELECTIONS AND WOULD HAVE SERIOUS IMPACT ON THE EDPE SELECTION PROCESS IN TERMS OF TIME AND EXPENSE.

THE AIR FORCE RESTATES ITS POSITION THAT 10 U.S.C. 2304 (G) IS NOT APPLICABLE TO "ORDERS PLACED UNDER CONTRACTS NEGOTIATED BY ANOTHER AGENCY, GSA," AND DENIES THAT "WHENEVER A DEFENSE AGENCY IS MAKING A SELECTION PRIOR TO PLACING AN ORDER UNDER ONE OF THESE CONTRACTS, THE PRICE, TERMS AND CONDITIONS OF WHICH HAVE ALREADY BEEN NEGOTIATED BY GSA, IT IS NEGOTIATING A CONTRACT PURSUANT TO 10 U.S.C. 2304 (G).' INSTEAD, THE CONTRACTS ARE NEGOTIATED BY GSA PURSUANT TO 41 U.S.C. 252. FINALLY, THE AIR FORCE SAYS THAT ALTHOUGH THIS IS NEITHER AN ADVERTISED PROCUREMENT UNDER 10 U.S.C. 2305, NOR A NEGOTIATED PROCUREMENT UNDER 10 U.S.C. 2304, ALTHOUGH SIMILAR IN SOME RESPECTS TO BOTH, POLICIES APPLICABLE TO BOTH ARE UTILIZED WHEN SUITABLE AND BENEFICIAL.

THE AIR FORCE ALSO DENIES AN "ARBITRARY SWITCH" IN ITS POSITION WITH REGARD TO ITS TECHNICAL ANALYSIS OF THE HONEYWELL SYSTEM. THE AIR FORCE MAINTAINS THAT IT HAS CONSISTENTLY ACKNOWLEDGED THAT IT DID NOT KNOW THE PRECISE REPAIRS NECESSARY TO MAKE THE HONEYWELL SYSTEM RESPONSIVE, BUT HAS IDENTIFIED AREAS WHICH ARE CONSIDERED LIKELY TO HAVE CAUSED THE FAILURE, WHICH, IN THEIR JUDGMENT, COULD NOT BE EASILY REPAIRED WITHIN A REASONABLE TIME.

THE AIR FORCE MAKES SEVERAL COMMENTS ON HONEYWELL'S ARGUMENTS CONCERNING THE "COST EFFECTIVENESS STUDY.' FIRST, IT ACKNOWLEDGESHONEYWELL'S STATEMENT THAT THE STUDY DID NOT CONSIDER CORRECTION OF THE TIMING DEGRADATION. IT DID NOT ASSUME ANY CORRECTION BECAUSE CONSIDERATION OF CLAIMS OF PERFORMANCE WOULD NOT BE APPROPRIATE IN FACE OF THE "HARD DATA" DERIVED FROM THE BENCHMARK. IT DEFENDS ITS EXTRAPOLATION OF COSTS OVER A 6-YEAR PERIOD, EVEN THOUGH THE COST GUARANTEE IS FOR ONLY 2 YEARS, AS THE IBM CAPACITY WILL SATISFY ITS ESTIMATED 6 YEAR LIFE DATA PROCESSING REQUIREMENTS, INCLUDING ANTICIPATED GROWTH, AND NO ADDITIONAL INCREMENTS OF EQUIPMENT WILL BE PURCHASED. THE REMARK ATTRIBUTED TO ASSISTANT SECRETARY MARKS CONCERNING THE SPECULATIVE NATURE OF THE STUDY IS CLARIFIED. THE TERM "SPECULATION" WAS USED IN REFERENCE TO THE COSTS ESTIMATED TO BE NECESSARY TO EXPAND HONEYWELL'S EQUIPMENT CAPACITY IN THE THIRD YEAR TO SATISFY THE ANTICIPATED GROWTH AT THE C AND D LEVELS. WITH REGARD TO HONEYWELL'S ARGUMENT THAT THE ANTICIPATED WORKLOAD GROWTH IS SPECULATIVE, THE REPORT INCLUDES THE FOLLOWING EXPLANATION:

THE FIRST PARAGRAPH ON PAGE 23 REFERS TO ANTICIPATED WORKLOAD GROWTH AS "ONE OF THESE SPECULATIONS.' ESTABLISHING DATA PROCESSING REQUIREMENTS ALWAYS NECESSITATES THE ESTIMATION OF PROBABILITIES. THE 200-HOUR MONTHLY OPERATIONAL USE REQUIREMENT CONTAINED IN THE RFP WAS INEXTRICABLY LINKED WITH AN ESTIMATED MINIMUM 10 PERCENT GROWTH RATE IN DATA PROCESSING REQUIREMENTS FROM THAT POINT THROUGH THE YEAR OF TRAINING, INSTALLATION AND SIX YEARS' LIFE, WHICH RESULTS IN THE SIXTH YEAR TYPICAL MONTH'S OPERATIONAL USE OF JUST ABOUT 400 HOURS OR THE 100 PERCENT EXPANSION ALSO CALLED FOR IN THE RFP. MOREOVER, THE 400 HOURS EQUATES TO 24 HOURS OF MANNING PER DAY, AND THE 10 PERCENT GROWTH RATE PER ANNUM IS BASED UPON EVIDENCE DRAWN FROM THE AIR FORCE'S PAST EXPERIENCE. THUS, THE AIR FORCE WAS ENTIRELY CONSISTENT IN EMPLOYING A 200-HOUR MANDATORY REQUIREMENT TIED TO AN ESTIMATED 10 PERCENT PER ANNUM GROWTH RATE.

THE AIR FORCE DENIES THE STATEMENT ATTRIBUTED TO SECRETARY BROWN THAT THE STUDY WAS NOT A "TRUE COST EFFECTIVENESS STUDY" AND STATES THAT IT ESTABLISHED THAT THE IBM PROPOSAL IS COST EFFECTIVE.

IBM WAS AFFORDED THE OPPORTUNITY TO EXPRESS ITS VIEWS IN THIS MATTER BECAUSE OF ITS VITAL INTEREST IN THE OUTCOME OF THE PROTEST. THIS IT DID IN A LETTER DATED JULY 6, 1967. MUCH OF THE LETTER RESPONDS TO THE ALLEGATIONS CONCERNING DELETION OF A MANDATORY REQUIREMENT FOR ITS BENEFIT AND THE RESPONSIVENESS OF ITS PROPOSAL TO ANOTHER MANDATORY REQUIREMENT. SINCE WE HAVE ALREADY STATED OUR CONCLUSION THAT THE AIR FORCE ANSWER TO THESE ALLEGATIONS WAS SATISFACTORY, WE WILL NOT RECOUNTIBM'S ARGUMENTS. THE THIRD MAJOR AREA DISCUSSED BY IBM CONCERNS THE APPLICABILITY OF 10 U.S.C. 2304 (G). IBM CONTENDS THAT THE STATUTE IS NOT APPLICABLE TO THE SELECTION OF A SOURCE TO PROVIDE DATA PROCESSING EQUIPMENT AND, ALTERNATIVELY, ASSUMING THE STATUTE DOES APPLY, THE AIR FORCE HAS COMPLIED WITH IT. FOUR REASONS ARE STATED IN SUPPORT OF THE PROPOSITION THAT THE STATUTE IS NOT APPLICABLE. THREE OF THESE ARE EXCEPTIONS WHICH ARE STATED IN THE STATUTE. FIRST, IS THE CONTENTION THAT THERE WAS "ADEQUATE COMPETITION" AND, IF IT IS CONCLUDED THAT THE STATUTORY REQUIREMENT FOR DISCUSSION HAS NOT BEEN MET, THE "ADEQUATE COMPETITION" EXCEPTION WOULD APPLY. SECOND, THE "ACCURATE PRIOR COST EXPERIENCE" EXCEPTION SHOULD APPLY TO OBVIATE THE NECESSITY FOR DISCUSSIONS BECAUSE OF THE AIR FORCE PAST EXPERIENCE WITH AT LEAST 30 PREVIOUS SIMILAR SELECTIONS AND BECAUSE THE PROCUREMENT INVOLVES STANDARD COMMERCIALLY AVAILABLE DATA PROCESSING EQUIPMENT. THIRD, DISCUSSIONS ARE NOT REQUIRED BECAUSE THIS IS NOT A PROCUREMENT "IN WHICH TIME OF DELIVERY WILL PERMIT" PROTRACTED NEGOTIATIONS AFTER SUBMISSION OF PROPOSALS. FINALLY, IBM SAYS THAT PUBLIC LAW 89-306 PRE-EMPTS THE APPLICABILITY OF 10 U.S.C. 2304 (G), AND ESTABLISHES, ALONG WITH BUREAU OF THE BUDGET CIRCULAR NO. A-54 AND DEPARTMENT OF DEFENSE DIRECTIVE 4105.55, THE APPLICABLE PROCUREMENT POLICY. ALSO CITED IN THIS CONNECTION IS ASPR 5-102.3. THE ALTERNATIVE ARGUMENT IS THAT THE REQUIREMENT FOR DISCUSSIONS WAS IN FACT MET BY THE EXCHANGE OF WRITTEN QUESTIONS AND ANSWERS WHICH PRECEDED SUBMISSION OF PROPOSALS, AND BY AN AMENDMENT HONEYWELL WAS ALLOWED TO MAKE AFTER SUBMISSION. IBM HAS ALSO INDICATED ITS CONCURRENCE IN THE AIR FORCE POSITION ON THE VALIDITY OF THE BENCHMARK METHOD OF VALIDATING PROPOSALS AND THE EVALUATION OF OVERALL COSTS ON A 6-YEAR BASIS TO JUSTIFY SELECTION OF ITS EQUIPMENT.

THE FOREGOING STATEMENT OF THE FACTS, CONTENTIONS, AND ARGUMENTS OF THE INTERESTED PARTIES HAS RAISED SUBSTANTIAL ISSUES CONCERNING VARIOUS TECHNICAL MATTERS, THE SOURCE SELECTION PROCEDURE, THE EFFECT OF VARIOUS PROVISIONS OF THE RFP, AND THE PROPRIETY OF THE ULTIMATE SELECTION OF IBM. HOWEVER, WE BELIEVE ALL OF THESE ISSUES ARE SUBSIDIARY TO, ALTHOUGH HAVING A BEARING ON, THE ISSUES ON WHICH THE HONEYWELL PROTEST REQUIRES OUR DECISION. THE FIRST SUCH ISSUE RAISES THE BASIC QUESTION WHETHER THE PROVISIONS OF 10 U.S.C. 2304 (G) ARE APPLICABLE TO THIS SOURCE SELECTION. THIS SECTION OF THE ARMED SERVICES PROCUREMENT ACT, WHICH WAS ENACTED AS AN AMENDMENT BY PUBLIC LAW 87-653, EFFECTIVE DECEMBER 1, 1962, PROVIDES AS FOLLOWS:

(G) IN ALL NEGOTIATED PROCUREMENTS IN EXCESS OF $2,500 IN WHICH RATES OR PRICES ARE NOT FIXED BY LAW OR REGULATION AND IN WHICH TIME OF DELIVERY WILL PERMIT, PROPOSALS SHALL BE SOLICITED FROM THE MAXIMUM NUMBER OF QUALIFIED SOURCES CONSISTENT WITH THE NATURE AND REQUIREMENTS OF THE SUPPLIES OR SERVICES TO BE PROCURED, AND WRITTEN OR ORAL DISCUSSIONS SHALL BE CONDUCTED WITH ALL RESPONSIBLE OFFERORS WHO SUBMIT PROPOSALS WITHIN A COMPETITIVE RANGE, PRICE, AND OTHER FACTORS CONSIDERED * * *.

WE BELIEVE THE ARGUMENT THAT THIS STATUTE DOES NOT APPLY BECAUSE THE AIR FORCE IS MERELY SELECTING EQUIPMENT WHICH WILL BE PURCHASED UNDER A GSA FEDERAL SUPPLY SCHEDULE CONTRACT, IGNORES THE REALITIES OF THE SITUATION AND IS CONTRARY TO THE INTENT OF THE POLICIES AND LAWS GOVERNING THE PROCUREMENT OF EDPE. WHILE PUBLIC LAW 89-306, APPROVED OCTOBER 30, 1965, WHICH HAS BEEN CITED IN SUPPORT OF THIS ARGUMENT, AUTHORIZES THE ADMINISTRATOR OF GSA TO COORDINATE AND PROVIDE FOR THE ECONOMIC AND EFFICIENT ACQUISITION OF EDPE, IT ALSO PROVIDES THAT THE RIGHT OF AGENCIES TO DETERMINE THEIR EDPE REQUIREMENTS, INCLUDING THE SELECTION OF THE TYPES AND CONFIGURATIONS OF EQUIPMENT NEEDED, WILL NOT BE IMPAIRED. IT DOES NOT EXEMPT SUCH SELECTIONS FROM THE LAWS AND REGULATIONS APPLICABLE TO PROCUREMENTS BY AGENCIES. IN THIS CONNECTION, IN ADDITION TO THE ABOVE- QUOTED SECTION 2304 (G), BOTH THE ASPR AND FPR INCLUDE PROVISIONS REQUIRING DISCUSSIONS WITH ALL RESPONSIBLE OFFERORS WHO SUBMIT PROPOSALS WITHIN A COMPETITIVE RANGE, PRICE AND OTHER FACTORS CONSIDERED. ASPR 3- 805 AND FPR 1.3-805. BUREAU OF THE BUDGET CIRCULAR NO. A-54, AND ITS IMPLEMENTATION IN DEPARTMENT OF DEFENSE DIRECTIVE NO. 4105.55, HAVE BEEN CITED BY THE AIR FORCE AS ESTABLISHING POLICIES ON THE SELECTION AND ACQUISITION OF EDPE. HOWEVER, PARAGRAPH 5 OF THE BOB CIRCULAR PROVIDES THAT ALL EDPE ACQUISITION TRANSACTIONS ARE SUBJECT TO PREVAILING POLICIES, LAWS AND REGULATIONS GOVERNING PROCUREMENT BY AGENCIES.

ALTHOUGH THE RFP PROVIDES THAT THE EQUIPMENT PROPOSED WILL BE PURCHASED OR LEASED FROM THE FSS, THIS FACT, IN OUR OPINION, IS NOT SUFFICIENT TO EXEMPT THE SOURCE SELECTION OF THE EQUIPMENT FROM THE REQUIREMENT FOR DISCUSSION WITH ALL OFFERORS WITHIN A COMPETITIVE RANGE. PARAGRAPH 1-15, ATTACHMENT I, TO THE RFP PROVIDES THAT THE CONTENTS OF THE PROPOSAL OF THE SUCCESSFUL OFFEROR WILL BE CONSIDERED CONTRACTUAL OBLIGATIONS. PARAGRAPH 1-16, ATTACHMENT I, RESERVES TO THE GOVERNMENT (THE AIR FORCE) THE RIGHT TO REJECT ANY AND ALL PROPOSALS, OR TO NEGOTIATE SEPARATELY WITH ANY SOURCE WHATSOEVER IN ANY MANNER NECESSARY TO SERVE THE BEST INTERESTS OF THE GOVERNMENT. PARAGRAPH 1-8, ATTACHMENT I, RECOGNIZES THE APPLICABILITY OF ASPR 3 506, CONCERNING LATE PROPOSALS RECEIVED BEFORE AWARD OF NEGOTIATED PROCUREMENTS. IN ADDITION, THE RFP PROVIDES FOR THE TRAINING OF AIR FORCE PERSONNEL BY THE SUCCESSFUL OFFEROR AFTER SELECTION IS ANNOUNCED, AND DELIVERY OF EQUIPMENT IS KEYED TO THE SELECTION DATE. WHILE GSA WILL EXECUTE THE CONTRACT BASED ON THE AIR FORCE SOURCE SELECTION, SELECTION OF THE SOURCE, UNDER CIRCUMSTANCES SUCH AS HERE INVOLVED, IS IN OUR OPINION A PART OF THE PROCUREMENT PROCESS AND SUBJECT TO THE REQUIREMENTS OF 10 U.S.C. 2304 (G) AND THE APPLICABLE PROVISIONS OF ASPR. IN ANY EVENT, WHETHER THE PROCUREMENT BE CONSIDERED AS BEING NEGOTIATED BY THE AIR FORCE OR GSA, THE FPR AND ASPR, AS ALREADY POINTED OUT, CONTAIN IDENTICAL PROVISIONS REQUIRING DISCUSSIONS WITH OFFERORS WITHIN A COMPETITIVE RANGE.

ASPR 3-805.1, CITED ABOVE, WHICH IS AN IMPLEMENTATION OF 10 U.S.C. 2304 (G), PRESCRIBES THE NEGOTIATION PROCEDURES TO BE APPLIED IN THE SELECTION OF OFFERORS FOR NEGOTIATION AND AWARD. THE PROVISION THEREOF WHICH IS PARTICULARLY FOR CONSIDERATION HERE IS AS FOLLOWS: 3-805.1 GENERAL.

(A) AFTER RECEIPT OF INITIAL PROPOSALS, WRITTEN OR ORAL DISCUSSIONS SHALL BE CONDUCTED WITH ALL RESPONSIBLE OFFERORS WHO SUBMIT PROPOSALS WITHIN A COMPETITIVE RANGE, PRICE AND OTHER FACTORS (INCLUDING TECHNICAL QUALITY WHERE TECHNICAL PROPOSALS ARE REQUESTED) CONSIDERED * * *. THE OBJECTIVE OF NEGOTIATION IS STATED IN ASPR 3-101, WHICH PROVIDES THAT WHEN PROCUREMENTS ARE UNDERTAKEN PURSUANT TO NEGOTIATION, MAXIMUM COMPETITION SHOULD BE OBTAINED TO THE END THAT THE PROCUREMENT WILL BE MADE TO THE "BEST ADVANTAGE OF THE GOVERNMENT, PRICE AND OTHER FACTORS CONSIDERED.' THE CONCEPT OF NEGOTIATION WAS DESCRIBED IN OUR DECISION REPORTED AT 45 COMP. GEN. 417, 427, AS FOLLOWS:

THE TERM "NEGOTIATION" GENERALLY IMPLIES A SERIES OF OFFERS AND COUNTEROFFERS UNTIL A MUTUALLY SATISFACTORY AGREEMENT IS CONCLUDED BY THE PARTIES. 10 U.S.C. 2304 (G) IMPLEMENTS AND CLARIFIES THE DEFINITION OF "NEGOTIATE" IN 10 U.S.C. 2302 (2) AND IT IS OUR VIEW THAT THE TERM "NEGOTIATE" MUST BE READ IN CONJUNCTION WITH 10 U.S.C. 2304 (G) TO INCLUDE THE SOLICITATION OF PROPOSALS AND THE CONDUCT OF WRITTEN OR ORAL DISCUSSIONS, WHEN REQUIRED, AS WELL AS THE MAKING AND ENTERING INTO A CONTRACT. * * *

WHETHER THE ABOVE STATUTORY AND REGULATORY REQUIREMENT FOR DISCUSSIONS WAS REASONABLY SATISFIED IN THE PROCEDURES EMPLOYED IN SELECTING IBMAS THE SOURCE FOR PHASE II IS THE SECOND ISSUE. IT IS OUR UNDERSTANDING THAT THE AIR FORCE SOURCE SELECTION PROCEDURE HAS SERVED THE DEPARTMENT WELL IN 30 PRIOR PROCUREMENTS AND IT IS REPORTED TO HAVE BEEN ACCEPTED BY THE COMPUTER INDUSTRY. THE ONLY ASPECT OF THE PROCEDURE WHICH HAS COME INTO QUESTION IN THE INSTANT CASE IS THE SO CALLED ,BENCHMARK" OR LIVE TEST DEMONSTRATION. WHILE THE USE OF THE BENCHMARK TO VALIDATE THE CAPABILITY OF A PROPOSED SYSTEM IS NOT AT ISSUE, THE FINALITY ACCORDED HONEYWELL'S FAILURE TO MEET THE BENCHMARK IN THE PRESENT CASE MUST BE JUDGED IN THE LIGHT OF THE REQUIREMENT FOR FURTHER DISCUSSIONS WITH ALL RESPONSIBLE OFFERORS WITHIN A COMPETITIVE RANGE, PRICE AND OTHER FACTORS CONSIDERED.

BOTH THE CITED STATUTE AND REGULATIONS REQUIRE SUCH DISCUSSIONS, WITH CERTAIN STATED EXCEPTIONS WHICH WE DO NOT BELIEVE APPLICABLE HERE. HAVE HELD THAT "COMPETITIVE RANGE" ENCOMPASSES BOTH PRICE AND TECHNICAL CAPABILITY. SEE 45 COMP. GEN. 417 AND B-159796, NOVEMBER 30, 1966. THE RESULTS OF A BENCHMARK TEST ARE THEREFORE SIGNIFICANT IN DETERMINING WHICH PROPOSERS SHOULD BE CONSIDERED IN A COMPETITIVE RANGE. CONCEIVABLY BECAUSE OF A DIFFERENCE IN PRICES, A COMPETITIVE RANGE MIGHT NOT INCLUDE ALL THOSE WHO PASSED THE TEST. ALSO, IT SEEMS TO US THAT FURTHER NEGOTIATIONS SHOULD NOT BE PRECLUDED EVEN THOUGH NO PROPOSER PASSED THE TEST, IF SEVERAL WERE CLOSE. THEREFORE, TO GIVE EFFECT TO THE STATUTORY AND REGULATORY REQUIREMENT FOR DISCUSSIONS AND FOR SUCH DISCUSSIONS TO BE MEANINGFUL, FAILURE TO PASS A BENCHMARK TEST SHOULD NOT AUTOMATICALLY PRECLUDE THE NECESSITY FOR FURTHER DISCUSSIONS.

WHEN THE APPLICATION OF A MANDATORY BENCHMARK TEST REQUIREMENT RESULTS, AS IN THIS CASE, IN LEAVING ONE PROPOSER, AND ITS PRICE IS, INITIALLY AT LEAST, SUBSTANTIALLY IN EXCESS OF THE PRICE OF ANOTHER PROPOSER WE BELIEVE THE SPIRIT AND INTENT OF 10 U.S.C. 2304 (G) WOULD NOT BE SERVED WITHOUT FURTHER DISCUSSION TO DETERMINE WHETHER THE OTHER PROPOSAL CAN BE IMPROVED TO MEET THE BENCHMARK REQUIREMENT.

CONCEDEDLY, BOTH IBM AND HONEYWELL ARE RESPONSIBLE OFFERORS. DISCUSSIONS WERE CONDUCTED, AT LEAST WITH HONEYWELL, AFTER PROPOSALS WERE SUBMITTED IN DECEMBER 1966. THE ONLY JUSTIFICATION FOR REFUSING TO CONTINUE DISCUSSIONS WITH HONEYWELL AFTER IT FAILED THE BENCHMARK TEST WOULD BE A DETERMINATION THAT ITS PROPOSAL WAS NOT WITHIN A COMPETITIVE RANGE, PRICE AND OTHER FACTORS CONSIDERED. WHEN THE BENCHMARK TEST LEFT ONLY ONE PROPOSER AS AN ELIGIBLE CONTRACTOR, WE DO NOT BELIEVE THE 200 HOUR BENCHMARK TEST, WHICH WAS THE ONLY TEST HONEYWELL FAILED, SHOULD HAVE BEEN CONSIDERED DETERMINATIVE OF WHAT CONSTITUTED A "COMPETITIVE" RANGE WITHOUT REGARD TO PRICE.

IT MUST BE RECOGNIZED THAT HONEYWELL HAS INSISTED SINCE SHORTLY AFTER IT WAS ADVISED OF ITS FAILURE TO PASS THE BENCHMARK TEST THAT IT COULD MEET THE 200 HOUR REQUIREMENT WITHIN A SHORT PERIOD OF TIME BY CERTAIN MODIFICATIONS OF ITS EQUIPMENT AT RELATIVELY LITTLE ADDITIONAL COST. COURSE THESE CLAIMS NEED NOT BE ACCEPTED WITHOUT SUBSTANTIATION. EVEN THOUGH THE AIR FORCE GIVES LITTLE CREDENCE TO THESE CLAIMS, THE AIR FORCE POSITION DOES NOT APPEAR SO CERTAIN THAT IT OBVIATES ANY REASONABLE POSSIBILITY THAT HONEYWELL COULD SUBSTANTIATE THEM. IT SEEMS TO US THAT THERE IS AT LEAST UNCERTAINTY WHETHER HONEYWELL CAN OR CANNOT MEET THE 200 HOUR REQUIREMENT WITHIN A REASONABLE TIME AND AT A RELATIVELY SMALL INCREASE IN PRICE IF AFFORDED THE OPPORTUNITY TO DO SO. FURTHERMORE, IF FURTHER DISCUSSIONS SHOULD RESULT IN A REDUCTION OF HONEYWELL'S TIME TO 200 HOURS, THE COST DIFFERENTIAL WOULD BE CONSIDERABLE. IT IS DIFFICULT TO UNDERSTAND, ON THE RECORD PRESENTED, WHY HONEYWELL WAS NOT AFFORDED THE OPPORTUNITY OF DISCUSSION AND EXPLANATION, PARTICULARLY IN VIEW OF THE POSSIBILITY THEREBY OF INCREASING COMPETITION, AND THE POTENTIAL SAVING INDICATED. THIS IS WHAT, IN OUR OPINION, THE CITED STATUTE AND REGULATIONS ENVISAGE AND REQUIRE.

IN VIEW OF ALL THE ASPECTS OF THE CASE, WE BELIEVE THE HONEYWELL PROPOSAL SHOULD HAVE BEEN CONSIDERED TO BE WITHIN A COMPETITIVE RANGE AND THAT FURTHER DISCUSSION WITH HONEYWELL SHOULD HAVE TAKEN PLACE BEFORE SOURCE SELECTION. UNDER THE CIRCUMSTANCES, FURTHER WRITTEN OR ORAL DISCUSSIONS SHOULD BE HELD WITH HONEYWELL AS WELL AS WITH OTHER OFFERORS WHOM YOU CONSIDER TO BE WITHIN A COMPETITIVE RANGE.

WITH REFERENCE TO THE QUESTION OF WHETHER THE SELECTION OF IBM AS A SOURCE WAS REASONABLE, WE CANNOT SAY THAT IT WAS BECAUSE OF THE FAILURE TO HAVE CONDUCTED FURTHER NEGOTIATIONS WITH HONEYWELL AFTER THE BENCHMARK TESTS. HOWEVER, EXCEPT FOR THIS FEATURE, WE ARE OF THE OPINION THAT THE PROCEDURE OF SOURCE SELECTION USED BY THE AIR FORCE IS A REASONABLE ONE. IT IS TO BE UNDERSTOOD THAT OUR OPINION AS TO THE REASONABLENESS OF THE PROCEDURE DOES NOT ENCOMPASS THE REQUIREMENTS OF THE AIR FORCE FOR EDPE NOR THE VALIDITY OF THE COST EFFECTIVENESS STUDY MADE IN THIS CASE.