Jim Carciofini

Research & Engineering

Constraint Based Reasoning, Hybrid Reasoning, Multi-agent Collaborative Systems


Jim Carciofini has performed research in computer science since 1984. Acting as the lead software architect on a wide range of software projects using a diverse set of technologies, he is sought out for his ability to analyze new problem domains and design and implement novel solutions.

Jim is the lead software architect for Galois’s Vulnerabilities out of Learning Automata (VOLTA) program that is funded under AFRL’s NOVA Program. VOLTA will provide new automated capabilities for performing vulnerability assessments of deployed embedded systems, and systems of systems.

Building on VOLTA, Jim is supporting Galois’s Detecting Emergent Computation In Multiple Automata Layers (DECIMAL) project, an 18-month effort, part of DARPA’s (AIMEE) program.

Jim is a contributor on DARPA’s Cyber Assured Systems Engineering (CASE) program, and on the previous DARPA GRAINS program, and leads the development of the mathematical optimization models in the medical logistics domain (medical logistics).

Jim was also the implementation lead for our Separation Platform for Integrating Complex Avionics (SPICA) program. He implemented an efficient mathematical optimization model (constraint-based) to find a schedule for the entire avionics system that meets complex timing properties and resource constraints. This work built on his previous work for Honeywell’s SAFEbus architecture (ARINC 659), which is still in use on aircraft today. The SPICA work has been extended to build an ARINC653 scheduler as part of the CAMET® Library, the Architecture Analysis and Design Language (AADL) tool suite Adventium (now Galois) has built for the Army Joint Multi-Role Technology Demonstration (JMR TD) program.

Jim was technical lead on Adventium’s Allocation of Missions Built on Resource Optimization (AMBORO) program. This program was part of DARPA’s Mission‐oriented Resilient Clouds (MRC) program.

Before his current work with Galois, Jim spent several years as the lead architect of Honeywell’s constraint-based software library. He was also a key developer using this set of tools to create novel solutions in several domains, including petroleum refinery recipe formulation and batch scheduling, batch manufacturing, continuous manufacturing, SAFEbus avionics processor and communications scheduling, satellite operations scheduling, and space shuttle operations planning.

Jim was the lead software architect on the Augmented Cognition program (AUGCOG), and the Neurotechnology for Intelligence Analysts program, two large DARPA research programs using real-time physiological data to augment and improve human performance on a variety of tasks.

He has been recognized with many awards in his service. In 1994 he won Honeywell’s highest engineering honor, the H.W. Sweatt Award, for his work on the SAFEbus avionics processor and communications scheduling tool for the Boeing 777. In 2006 he was recognized with the Honeywell Aerospace Technical Achievement Award for his work the DARPA Augmented Cognition (AugCog) program. In 2009 he was again awarded the Honeywell Aerospace Technical Achievement Award as a member of the team on the DARPA Neurotechnology for Intelligence Analysts (NIA) program.
Jim holds a B.S. in Computer Science from the University of Minnesota.