The Rapid Autonomy Integration Lab (RAIL) is an effort out of the service’s unmanned maritime systems program office… “The vision for the RAIL is that many industry partners … will bring their code to the RAIL,” Small said. Through testing and standardized tools, the government and industry “will integrate those software into a holistic baseline that can then be tested and certified and deployed to a platform.”
… The management role, known as the “autonomy baseline manager,” is intended to be “that support infrastructure to help manage the process, enforce the standards, configuration management, and just overall execution of the development and certification period and software.”
That was from a Breaking Defense article. This seems to be a different approach than the Army and Air Force are taking with their autonomy efforts. The Air Force is using a government reference architecture from AFRL and outsourced Skyborg’s autonomy core system to Leidos to integrate the government’s baseline. The Army is acting perhaps stronger than the Air Force in terms of being the lead systems integrator.
From the outside, the Navy seems to be taking a much more federated development process for its unmanned surface and underwater vessels, the configuration of which will be managed by an autonomy baseline manager. To some degree, this feels a bit like the Navy’s heritage with the Polaris missile development where there was a number of parallel subsystem developments managed by the Navy but supported by Ramo-Woolridge as a non-performing lead systems integrator.
I wonder how different the “baseline manager” will be from the functions of a “lead systems integrator.” About 15 years ago, however, Congress limited DoD’s use of outsourced lead systems integrators:
Except as provided in subsection (b), no entity performing lead system integrator functions in the acquisition of a major system by the Department of Defense may have any direct financial interest in the development or construction of any individual system or element of any system of systems.
The exceptions are when the SecDef certifies that the LSI is (1) constructing a significant portion of the system; and (2) was selected through a competitive process.
A lead systems integrator, by the way, is defined as:
(A) a prime contractor for the development or production of a major system, if the prime contractor is not expected at the time of award to perform a substantial portion of the work on the system and the major subsystems; or
(B) a prime contractor under a contract for the procurement of services the primary purpose of which is to perform acquisition functions closely associated with inherently governmental functions with respect to the development or production of a major system.
Usually this limitation wouldn’t affect most defense programs since the prime often does a “significant” portion of the work including assembly or otherwise. However, for the Navy RAIL effort, it would seem that the contractor will largely be doing configuration management and testing. The Navy may consider it to be doing the real work of lead systems integrator. But it will be interesting to see how the services different approaches to integrating autonomy work out.
One big issue here is that many rules and regulations apply to well-defined programs, whereas the autonomy efforts are more of a suite of interrelated capabilities. The Army’s Future Combat Systems was a networked system of ground vehicles which failed miserably, leading to the laws against outsourced LSIs. For autonomous systems, there will likely be a number of individual programs worked on by different organizations throughout the Navy.
Leave a Reply