Project Turnaround

After completing ISDN BRI testing, I was assigned as a contractor to be the test department technical lead for ISDN PRI. Prior to my taking on the role, a decision had already been made to purchase an off-the-shelf stand alone product from a Canadian company and use it for the DMS 10 ISDN PRI offering. The primary customer being GTE (which later became Verizon).

The contract with GTE was for both ISDN BRI and ISDN PRI. The delivery of ISDN BRI had already been seriously delayed. Fortunately, GTE waived the penalty but said ISDN PRI had to be delivered on time, or penalties reportedly in the multiple millions of dollars would be pursued.

As we began design meetings to determine what development was needed on the DMS 10 to accomodate this black box, a general belief existed that the PRI schedule was safe but this belief was tempered by frequent use of doubtful words when referring to the converter box.

However, as I began reading the SS7 specification, things weren't adding up. The SS7 spec was new to me but I really knew PRI well. What I was seeing simply didn't add up. There was no way that some information elements and even entire messages could be mapped from PRI <-> SS7 without the use of ISUP -- which the box didn't support.

My design counterpart, also a contractor who I had previously worked with at Siemens agreed that he couldn't see how this box would ever work. We prevailed upon management to have the vendor send us a box to evaluate. The vendor balked but finally agreed to send a box and an engineer down who by provisions of the NDA would essentially be the only person permitted to touch the box, or administer it. That was fine as we had no desire to look inside. Once the NDA was signed, an engineer flew down from Canada along with the PRI box.

Once we set everything up in the lab, the basic protocol tests we ran against the box began revealing serious flaws. My development counterpart, brought our test results to his manager who passed them higher. Not long after that, we knew serious consideration was being given by upper management to our test report because requests were going out for development estimates. In due time, a decision was made to abandon the black box approach and to pursue a fully in-house developed product.

The decision to abandon the black box was not trivial because millions of dollers were at stake should our delivery schedule slip. Since roughly 25 developers and testers would be needed for an in-house effort, the internal project risks were considerable in light of the schedule. Never-the-less, the decision was that an existing T1 board would be used for layer 1. A layer 2 Q921 protocol stack would be purchased and modified as necessary. ISDN PRI Layer 3 would be fully developed in-house.

With the black box behind us, we were ready to dive into the in-house development of ISDN PRI... an effort expected to generate significant revenue.