Autonomous Systems in Defense – Robert’s Rules for Robots: Part 2

21 November 2018
vehicle.jpg

In the first part of this two part post, I noted how the autonomous systems that are central to the military’s future will depend extensively on AI, machine learning and deep learning via neural networks. The post concluded with a reference to two autonomy research programs launched by DARPA.

The first of these is Explainable AI (XAI) This program aims to create a suite of machine learning techniques that:

The second program is Assured AI. Continual assurance is defined as an assurance of the safety and functional correctness of the system provided provisionally at design time, and continually monitored, updated, and evaluated at operation-time as the system and its environment evolves. https://www.darpa.mil/program/assured-autonomy

The concept and reality of human-machine teaming is also being explored in very tangible ways, as evidenced by the recently-opened Air Force Research Laboratory University Center of Excellence at Carnegie Mellon, where psychology and the digital domain intersect.

Vast computational power

As the research into collaborative autonomy unfolds, there is also recognition that the computational power required to run AI will be vast. Intel, for example has projected bringing to market processors capable of 50 to 100 TeraFLOPS by 2030. These AI-capable HPEC systems in many cases will take the form of small form factor multi-domain controllers built into compact platforms - all the while generating kilowatts of heat.

Abaco’s capabilities in thermal management, ruggedizing and SWaP will provide enduring value in this emerging market. Additionally, maintaining the close relationships we enjoy with our silicon partners will help ensure we continuously plan the right products around the most advanced processors.

Many customer meetings and defense symposia have highlighted the importance of AI and autonomous systems, and numerous RFIs and RFPs we receive include autonomous capabilities and where Abaco can play an important role.

The Intel Federal Summit only served to confirm the growing importance of AI, with keynotes, round tables, exhibits and customer breakouts featuring the ongoing developments in AI, machine learning and deep learning. Many visitors to our tabletop in the Technology Showcase were directly involved in autonomous systems and understood the need for rugged as these systems move from the bench into the real world.

There are exciting, if challenging, times for the military – and for Abaco. You might want to read our further thoughts on the subject in this white paper.

Kevin Moran

Kevin joined Abaco Systems in January of 2017 as Business Development Director focused on the autonomous market. He has over 20 years of experience in intelligent and connected vehicles and systems, with an engineering degree from the University of Michigan.