The program imposed an extensive, though rational, set of SDRL requirements that reflected the Government's need to assure the correct operation and ongoing supportability of the design. It was determined that the basic COTS system design and documentation would be the primary mechanism used to meet the SDRL requirements. The only exceptions to this were the Requirements Traceability Matrix, the Acceptance Test Document and the Intellectual Property Report. These documents were exclusive to the ORTT contract.

It was vitally important for the success of the project that development and productionizing of most of these sub-system elements could proceed completely independently of the overall program. Most importantly, these developments were not contingent upon design data, agreement on the functions, and then the actual successful production of the replica panels. In the actual event, this proved to be an extremely important consideration since there are many and various woeful stories to be told regarding the manufacture of these panels. Many companies were found wanting, and even those that met some of the initial test evaluations proved incapable of maintaining their quality of performance. (The value and importance of establishing collaborative relationships with trusted vendors was once again proved. There is a presumption today that everything is a "quality-less" commodity; that success can only be achieved by depriving the other member of the transaction of a fair profit. The reality is that efficiency and quality demands long term mutuality of profit.)

Since each of these sub-system elements was thoroughly tested as a stand-alone production item, the problems encountered in testing the finished panels were usually restricted to cabling and connection issues that were resolved during first-article testing.

The final production quantities of equipment were received just a few weeks (and, in some cases, days) prior to full-up system testing. Errors and discrepancies found in final system test were resolved in the following few weeks. Absent these minor problems, we would have delivered a month ahead of schedule.

The schedule did not support testing of all the combinations and permutations of system operation. For this to have been carried out rigorously would have required squads of uniformed SME's to man all the positions and a script worthy of Cecil B. de Mille.

The four day acceptance testing schedule was as follows:

1. Formal PCA
2. Formal Cold Start
3. FCA

This was conducted against a formal RTM based ATP. The process was restricted to inspecting that each device type met the requirement; and that factory testing records supported the assumption that testing one device type equaled testing all devices.

It also ensured that the sound quality was good, that the state machine logic was correct, and established that each position was connected into the environment correctly. (The completeness and adequacy of the Model Builder radio/communications environment was a COTS predicate and established at previous PDR/CDR demonstrations.)

The confidence that enabled such a limited testing scenario was due to several factors: the credibility that had been established by ASTi's performance over the course of the program, the fact that the customer was allowed an unlimited free-play session prior to sign-off, the two years design and hardware warranty, the generous withhold against acceptance at the customer site in Montreal.

(In fact, it is a general operating principle at ASTi to insist upon a substantial withhold against final on-site satisfaction, as it greatly reduces the non-value added activities associated with extensive acceptance testing.)

For a small company such as ASTi to provoke anxiety and the consequent intense scrutiny of a major corporation such as Lockheed Martin on a major training program like the ORTT (which had already had its fair share of media and parliamentary prominence) was considered to be a recipe for disaster. The crux of the game plan was to stay ahead of the power curve in all respects. To submit all SDRLS on-time, if not early. To resolve all issues before they were due, and to present not a design, but a working prototype at the design reviews.

Some spirited arguments at early stages of the program established the concept's underlying COTS deliverables, and allowed us to avoid inappropriate levels of documentation, such as circuit wiring diagrams, internal assembly drawings, or fault finding below the recommended LRU level.

The existing O&M Manuals were the basis of all submissions. For new or changed sections, revised indexes were provided to maintain a high comfort-level that the desired documentation would eventually be forthcoming.

The ORTT system was designed, assembled, integrated, tested, delivered, and installed at the customer's facility in less than 15 months.

Our success with the ORTT communications simulation subsystem was a direct result of our pre-existing communications simulation capability combined with our ORTT system approach incorporating:

• Distributed processing resources
• Separate control and audio networks
• Electrically identical panel units
• Use of 100 Mb ethernet interface
• Model Server architecture for centralized panel development
• Extensive use of software configuration files
• Purchasing the final computing platforms at the latest possible moment to maximize bang-for-the-buck
• A comprehensive library of standard documentation which was the basis for virtually all the technical SDRL submissions.
• The pre-existing capability of the DACS to handle heavily loaded networks and distribute large volumes of audio across an internal (DIS) network.

Next page: Detailed System Requirements