SRC’s machine intelligence and autonomy capabilities enable sensing and understanding, strategy and action selection, resource management, and autonomy to provide optimal performance under complex, dynamic conditions.

We are working to identify, incubate and accelerate the development of machine learning (ML), artificial intelligence (AI), and autonomous system capabilities for our products and services to help warfighters maintain superiority in today’s complex battlespace.

SRC has streamlined machine intelligence and autonomy capability development and integration into mission solutions — while maintaining the agility needed to provide timely solutions to critical customer challenges. Our engineers utilize a multidisciplinary development approach that embraces both diversity of perspective and talent to develop innovative solutions for our customers.

SRC focuses on five main areas to develop innovative machine intelligence and autonomy capabilities:

• Education and Internal Growth: SRC has developed an advanced internal curricula for machine learning and artificial intelligence that includes knowledge discovery and sharing as well as contests and competitions to promote innovation and ideation.
• Infrastructure: We leverage large scale ML Ops/DevOps development ecosystems, traditional and container-managed high-performance computing systems as well as data, code and pipeline curation, discovery, and management techniques to quickly and securely develop machine intelligence capabilities.
• Data, World Models and Training Environments: SRC leverages advanced multi-domain data generation, synthesis and growth for machine intelligence and autonomous system training and evaluation. Using these capabilities, we have developed integrated live, virtual and constructive (LVC) simulation tools for scenario generation and adversarial training.
• Modular Algorithms: We promote cross-enterprise development of “best of breed” deep model structures and have created processes to align our development, training, and deployment pipelines. Shared heterogeneous knowledge repositories and testbeds along with our development processes support rapid integration of machine intelligence subsystems into comprehensive mission systems and customer solutions.
• Multi-Purpose Hardware: We design and develop machine intelligence-enabled hardware solutions, integrating heterogeneous processing components (GPPs, GPUs, DSPs, FPGAs, and neuromorphic computing techniques) into rugged and highly-constrained environments. We also develop modular open architecture machine intelligence hardware sub-systems, and machine intelligence development and evaluations tools.