CDR Joseph Cohn, PhD ONR Code 341 Division Deputy

Transcription

1 CDR Joseph Cohn, PhD ONR Code 341 Division Deputy

2 Disclosure Information 84th Annual AsMA Scientific Meeting CDR Joseph Cohn I have no financial relationships to disclose. I will not discuss off-label use and/or investigational use in my presentation

3 Pre Decisional Unmanned Aircraft Systems Evolving Role of Human in Aviation Systems Human Role in Flight Environment Managerial Directive Aeromedical Physical Visual Acuity Muscle Balance Direct interaction <WW I Aeronautical Ability Intelligence Psychomotor Spatial Reasoning Blend direct and indirect interaction WW II Time Aviation Selection Test Battery Psychomotor Technical Acumen Personality Human gets a vote 1990s 2000s UAS Shift towards cognitive / perceptual tasks Information-rich, distributed, lengthy, collaborative missions Increasing reliance on Automation & Multitasking Today What are the implications for selecting, training and equipping these Air Warriors? FOUO 3

4 Some Important Questions Common Control Station (CCS) POR Memo: Multiple UAS platforms will integrate into CCS How to enable effective information display for safe and effective UAS operations? SECNAV Unmanned Systems Goal: Supporting a Human Capital &Training Strategy for UxS How to provide optimally selected, effectively trained cadre of UAS operators? N2/N6 Information Dominance Roadmap for UxS: How to optimize UAS DOTMLPF-P? Navy Aviation Simulation Master Plan: Almost all UAS training must be conducted in simulated environments - How to provide the agent-technologies to populate scenarios for effective training? Answers require an integrated human system solution 4

5 USMC Group 4 MQ-9 UCLASS % Accidents Pre Decisional -DoD UAS Roadmap Williams, 2004; see also Tavaryanas, 2006; Herz, 2008 UAS UAS capabilities capabilities will will grow grow in in complexity complexity DoN DoN will will soon soon establish establish aa significant significant UAS UAS Inventory Inventory - From Aircraft Investment Plan FY , submitted with the FY 11 Defense Budget historically historically up up to to 50% 50% of of UAS mishaps attributed to UAS mishaps attributed to human human factors causes factors causes change, change, but but frequency frequency remains remains consistent consistent Today, building UAS with the requirement that the Operator adapt to the System FOUO 5

6 The Payoff of a Human Systems View Done in parallel What makes an effective UAS Operator? Operational Safety & Effectiveness SELECTION 1. Select based on KSAs 2. Up front costs (600 operators, FY20) a. Military Aviators: Basic + Platform-specific training >$1M b. Non Military Aviators: $392M avoided by NOT paying for manned flight training 1. Common Control System SDD calls for controlling: Multiple UAS; Dissimilar type; 1 operator 2. Enabling fewer operators to safely and effectively control multiple UASs by understanding how to support KSAs through display technologies` DESIGN TRAINING 1. Target relevant KSAs 2. Through modeling & simulation, enable access & availability of training on Common Control System 6

7 Selection Operational Challenge Optimizing Manpower/Personnel How we select today Wide range of UAS missions What are UAS KSAs? Different from Manned Aviation Works for Manned Aviation Proper selection saves training $$ and should reduce mishap rates too AQR and SNA Attrition AQR Scores vs Attrition AQR Scores for Field Flight Performance Board Attendees 45% 40% Saves $38 Million in training costs/yr % Attrition % Attrition 35% 30% 25% USN 20% USMC 15% 10% 5% 0% AQRAQR Leverage from manned aviation; create anew through experimentation Joseph.cohn@navy.mil 7

8 Training Operational Challenge Making it easier to develop content Naval Aviation Simulation Master Plan: Almost all UAS training must be conducted in simulated environments Need: 1,000 s of entities to populate training scenarios Take Sensor Data Extract Key Features Content Development: Requires significant expertise hrs for 1 content hr Classify/Organize (AI/ML) Model (Computational Cognitive Models) Next Steps: Automate generative SAF behavior development Integrate into Navy Training Framework Take advantage of rich data sources from UAS & advances in cognitive modeling 8

9 Interface Operational Challenge Effectively Presenting Information What the future holds for UAS control well known Control Transfer/Hand off Multi Vehicle Control Autonomy VS What we (still) have to meet the future with - Information Dominance Roadmap - DoD UxS 2011 (unsigned) Workload Management Automation Alone is Not the Solution Experiment-driven designs to maximize UAS operator information processing Joseph.cohn@navy.mil 9

10 Common Control Station A Window of Opportunity Initial content for training UAS Sensors Battlespace data Make sensor information understandable Common Control Station Common Control Station Via Presentation Layer, Via Presentation Layer, Operators Window into UAS battlespace Operators Window into UAS battlespace (Live or Simulated) (Live or Simulated) Outcome Measures UAS Training Architecture Simulated training battlespace Generate realistic synthetic entities from raw sensor data UAS Operators Select UAS personnel using virtual missions 10

11 Pre Decisional Addressing the Challenges: A Path Forward UAS Operator Selection Identify the knowledge skills and abilities relating to operating Navy/USMC UAS platforms, according to platform and mission Identify and develop the specific UAS selection tests, and develop data collection instruments to capture test measures Identify the costs/benefits of using different candidate AVO populations UAS Operator Training Develop the knowledge structures needed to capture source data Define boundaries of behavior patterns of interest. Develop representative cognitive models from behavior data Use machine learning techniques to generate novel, doctrinally accurate, SAF behaviors from the existing modeled behaviors UAS Operator Interface Identify the core information requirements, platform common and platform specific, leading to successful operation of Group 4 and 5 UAS platforms Apply human factors and cognitive engineering principles to develop display design options that address identified information requirements Assess key AVO performance characteristics including: situational awareness, workload, vigilance and related indicators of cognitive performance, as well as related psychomotor measures 11

12 Parting Thoughts We ve been down this road before Commonality requirement provides a unique opportunity to get it right Keep an open mind Manned vs Unmanned Aviation culture may suggest different approaches than the data suggest UASs are becoming an integral part of our Operational environment Require as serious consideration as manned systems 12