FLA2SH: Flexible Levels of Adaptable Autonomy for Sensor Handling

ABSTRACT: SIFT, with subject matter expertise from the UAS pilots and sensor operators (SOs) at UAV Associates and in sensor processing capabilities from Ball Aerospace, will develop FLA2SH (Flexible Levels of Adaptable Autonomy for Sensor Handling) to extend and integrate with AFRL"s successful Flexible Levels of Execution-Interface Technologies (FLEX-IT) architecture for multi-modal delegation interactions with adaptable automation. FLA2SH will apply flexible delegation concepts to the functions of UAS SOs and to the novel problem of delegation for teams of SOs and pilots. We identify SO needs, then develop display and control concepts built around the hierarchical control methods available in FLEX-IT. Several needs are identified and initial concepts generated in the proposal itself, including display approaches, widgets and verbal interactions for fine-grained sensor controls, means to address loss of context and context switching, etc. Techniques for applying adaptable automation to sensor data processing itself are proposed as well. Concepts will be vetted for feasibility with sensor processing experts from Ball Aerospace. Phase 1 will produce sketches and storyboards for multiple concepts, will implement and demonstrate some of them, and will review all concepts with a panel of UAS operators for feedback and prioritization, leading to revisions and experimentation in Phase 2. BENEFIT: FLA2SH will open a new front in human-automation collaboration for multi-UAS opera-tions: effective, practical, controllable sensor operations. Sensor processing is the primary bottleneck restricting multi-UAS operations today, so successful results will break down a significant barrier to that goal. Since sensor data is scaling up far more rapidly than the number of UASs, SO support is arguably more necessary than pilot support. Improvements in the usability of multiple sensors from multiple UASs will improve the success rate of most UAS missions, improve operator awareness, while concurrently reducing the manning, training and attentional demands of operating multiple sensors and/or increasing the manageable numbers of vehicles. While many potential markets for such approaches exist across a wide range of applications and domains, we estimate the size of the market for FLA2SH to support military and commercial UAS control stations alone at about $3B annually in 2020, with another $.8B in precision agriculture business. We will pursue transitioning FLA2SH through a mix of licensing an associated toolkit and providing services to construct specific tools using it ourselves. The core science of FLA2SH can inform and benefit a much wider market, however, including human-automation inter-actions in medical, transportation, manufacturing and energy production systems.