Applying the sidekick principles to the prototyping of human-machine teaming in collection planning

Abstract

Collection planning for multi-domain operations is a complex task that is heavily manual, resource-intensive, and sub-optimally setup for collaboration. To solve these gaps and leverage advances in Artificial Intelligence and Machine Learning, which suggest that those are resolvable, our team employed the sidekick principles for human-machine teaming to design and prototype COGSWORTH, a system for Collection Optimization and Generation Synchronized with Optempo Reasoning Tailored to Humans. COGSWORTH is envisioned as a cognitive assistant to help collection planners. We report here on (1) our application of the sidekick principles to the prototyping of COGSWORTH, (2) how the sidekick principles are tangibly reflected in prototype artifacts, and (3) lessons learned and practical takeaways that may inform other researchers and practitioners in the human-machine teaming field.