By Dean ~ August 19th, 2011. Filed under: SDR & SCA.
“The dramatic scaling back of a U.S. Army radio program will save the Pentagon $15 billion in projected costs, the U.S. Defense Department told members of Congress.
Army officials told lawmakers in May that the Joint Tactical Radio System Ground Mobile Radio (JTRS GMR) would breach congressional spending limits under the Nunn-McCurdy statute. The service reduced planned buys of the system from 86,956 to 11,030, prompting a critical Nunn-McCurdy breach, according to Aug. 10 information paper released by DoD on Aug. 12.
A Government Accountability Office report found costs for the program’s research and development had grown by 69 percent from 2002 to 2011.”
I’ll bet that was an uncomfortable meeting with Congress! In any event, I object to the use of the word “save” here. The money already invested into this program has been wasted. They’re just not pouring more good money after bad.
So, what went wrong? The Wikipedia page (currently) carries this gem of an explanation:
“The JTRS program has been beset by delays and cost overruns, particularly Ground Mobile Radios (GMR), run by Boeing. Problems included a decentralized management structure, changing requirements, and unexpected technical difficulties that increased size and weight goals that made it harder to add the required waveforms.”
At this point, I should share a few disclaimers. First, I have no personal / direct knowledge of the internal issues with this program. Second, Foresight Systems M&S was not involved with the GMR program (this is actually one of the problems, but I’m jumping ahead of myself). Finally, the issues with the GMR program do not necessarily carry over to any of the other JTRS programs. As I’ve discussed in previous posts, programs like HMS are making good progress.
From what I’ve heard in “over beers” type discussions, the GMR program was plagued by ineffective mapping of the program requirements to the proposed platform(s), and unwillingness to negotiate realistic requirements. When confronted with these challenges, it seems that the default response was to add more computing power to the platform, without accounting for the added costs in terms of weight, power consumption, heat dissipation, complexity, reliability, financial impact etc.
Whatever the “official” explanation, I believe that it is largely a failure of system engineering. Our job as engineers is to work with the requirements to design the system that optimally achieves the requirements and meets the given physical / schedule / budgetary constraints. It’s also our job to push back, with evidence, on requirements that are unrealistic. It is most assuredly not our job to keep throwing hardware at the problem until we smell smoke.
If the parties involved in the GMR program had used a solid Performance Engineering methodology from the start, they would have been able to model the performance of various design alternatives. This would have enabled them to develop a solid understanding of the tradeoffs they would be making. The design team could have then been confident that the solution was fully optimized across all of the constraints and requirements before committing to a physical implementation. Instead, they seem to have made a bunch of incorrect assumptions and plunged ahead with implementation. The results speak for themselves.
Disciplined application of our Performance Engineering solutions will prevent these sorts of money, time and security wasting failures.
Of course, this isn’t bad news for everyone. Again, from the Wikipedia page:
“The U.S. military no longer plans to quickly replace all of its 750,000 tactical radios. The program is budgeted at $6.8 billion to produce 180,000 radios, an average cost per radio of $37,700. Program delays forced DOD to spend an estimated $11 billion to buy more existing tactical radios, such as the U.S. Marine Corps’ Integrated, Intra-Squad Radio, the AN/PRC-117F and the AN/PRC-150.”
Loyal readers of this blog might recognize that the AN/PRC-117F(C) and AN/PRC-150 are products from Harris Tactical Communications that are already in production and apparently selling well, very well.
My point is straight forward, even if I took a bit of a round about path to get there. You simply cannot produce competitive products based on complex, highly constrained, embedded systems (like military grade software defined radios) without rigorous Performance Engineering throughout the design process. We know this. We’ve had a long history with JTRS. We work on these systems every day. Learn from the failure of the JTRS GMR program and use Performance Engineering for your designs.
Technorati Tags: System Design, modeling, simulation, Model Driven Design, Model Based Systems Engineering, Model Driven Systems Engineering, Joint Tactical Radio System, Software Defined Radio, SDR, JTRS