My last endeavor before retiring was to sell and manage the $16M 5yr development of an advanced prototype surveillance sensor for shipboard use. The goals were emissions-free search of the open ocean horizon using infrared wavelengths to find low-flying missiles, and also when near shore to be able to monitor small boats. The three main hardware components were: 1) a highly specialized and unique $5M infrared focal plane array, 2) custom $1M wide field of view anamorphic optics, and 3) a $.8M stabilization system.
Besides these contracts, other tasking was accomplished with a small number of qualified people, especially for digital electronics interfacing with software, and there were no committees. The Office of Naval Research wanted to "own" the sensor, so at the Naval Research Labroatory we took responsibility for both hardware integration and software development. I wrote the three involved and difficult hardware specifications with adept critique from my boss. And I especially sought out, studied and avoided mistakes of other electro-optical development programs of the era.
I did make two mistakes, though. In the stabilization RFP (Request For Proposal) I removed an instruction to use "knuckle-joint" design. And in debate with ONR about the sponsor's desire to invest in development of a laser confirmation sensor, I was too slow to argue that the requirement for missile detection range by a laser should match the maximum possible, not the minimum required detection range by the horizon infrared search sensor. The basic sensor performed quite well, but my mistakes meant it had to be demonstrated on a non-optimal and bulky gimbal stabilization turret.
Navy Distributed Aperture Sensor Infrared Search and Track System Overview Waterman2004DAS.pdf