Air Force Office of Scientific Research Dynamics

Air Force Office of Scientific Research Dynamics & Control Program Overview Lt. Col Scott Wells, Ph. D Program Manager AFOSR/ND Air Force Research Laboratory

Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary • • AFOSR Overview Portfolio Overview – Management Summary – “Technical” Summary Future Direction/ Ideas Conclusions • • Future Direction/ Ideas Conclusions 2

AFOSR Mission Introduction AFOSR Overview Expand the horizon of scientific knowledge through leadership and management of the Air Force’s basic research program by investing in basic research efforts in relevant scientific areas. Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions Creating revolutionary scientific breakthroughs for the Air Force Central to AFOSR’s strategy is the transfer of the fruits of basic research to industry, the academic community, and to the other technical directorates of AFRL. Check out www. afosr. af. mil 3

Organization Air Force Office of Scientific Research Introduction DIRECTOR AFOSR Overview Dr. Brendan Godfrey Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas CHIEF SCIENTIST DEPUTY DIRECTOR Dr. Thomas W. Hussey SENIOR RESIVIST Lt. Col. Joe Fraundorfer Col. Michael Hatfield Conclusions AEROSPACE & MATERIALS SCIENCES Dr. Thomas Russell Dr. Don Carrick INTERNATIONAL OFFICE Dr. Mark Maurice HUMAN RESOURCES Ms. Terry Hodges MATHEMATICS, INFORMATION & LIFE SCIENCES ASIAN OFFICE OF AEROSPACE R&D PHYSICS AND ELECTRONICS EUROPEAN OFFICE OF AEROSPACE R&D Dr. Ken Goretta STAFF JUDGE ADVOCATE DIRECTORATE OF POLICY AND INTERGRATION Maj. Michael Greene Maj Ryan Umstattd Dr. Genevieve Haddad Col. Stephen Pluntze DIRECTOR OF CONTRACTING Ms. Trish Voss 4

AFOSR Research Areas Introduction Aerospace & Materials Sciences AFOSR Overview Physics & Electronics Mathematics, Information & Life Sciences Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions • • • Structural Mechanics Materials Chemistry Fluid Mechanics Propulsion • • Physics Electronics Space Sciences Applied Math • • Info Sciences Human Cognition Mathematics Bio Sciences Areas of Enhanced Emphasis - Information Sciences - Mixed-Initiative Decision Making - Adversarial Behavior Modeling - Novel Energy Technology - Micro Air Vehicles - Nanotechnology 5

AFOSR Supports Basic Research Introduction Basic Research Funding Foster Revolutionary Basic Research 728 Research grants at 211 universities AFOSR Overview Portfolio Overview AFOSR 194 Research projects at AFRL • Management Summary • Technical Summary 186 STTR contracts Future Direction/ Ideas Build Relationships Conclusions 39 Postdocs at AFRL 90 Summer Faculty at AFRL 37 Personnel exchanges Asian Office of Aerospace Research and Development 264 Short-term foreign visitors 58 Technical workshops European Office of Aerospace Research and Development Southern Office of Aerospace Research and Development 205 Conferences sponsored New in May 07 6

FY 07 AF Core Basic Research Investment By Discipline Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions 7

PORTFOLIO OVERVIEW Dynamics & Controls Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions NAME: Scott Wells YEARS AS AFOSR PM: 8 months BRIEF DESCRIPTION OF PORTFOLIO Developing theory, algorithms, and tools for reliable, practical design and analysis of high performance robust and adaptive control laws for future AF systems operating in uncertain, complex, and adversarial environments SUB-AREAS IN PORTFOLIO Control and dynamics of Unmanned Aerial Vehicles (single/multiple agent) • Autonomous Single Agent/Enabling Technologies • Cooperative Multiple Agent Aerodynamic flow control and control of unsteady phenomena Active waveform control Dynamics and Modeling (modeling, identification and uncertainty characterization) General control theory (nonlinear, adaptive, hybrid) Validation & Verification (V&V) 8

PORTFOLIO OVERVIEW Dynamics & Controls Introduction AFOSR Overview Sub-Area Distribution (Includes FY 06 & FY 07 Projects) Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas us mo ono Aut UAV Conclusions Co ope r UA ative V 9

PORTFOLIO OVERVIEW Dynamics & Controls Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions 10

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview • – – Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions Control and dynamics of Unmanned Aerial Vehicles (single/multiple agent) Autonomous Single Agent/Enabling Technologies Cooperative Multiple Agent • • • Aerodynamic flow control and control of unsteady phenomena • • General control theory (nonlinear, adaptive, hybrid) Active waveform control Dynamics and Modeling (modeling, identification and uncertainty characterization) Validation & Verification (V&V) 11

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Unmanned Aerial Vehicles (UAV) • Dynamics & Autonomous UAV Control • Cooperative Multi-agent Dynamics and Control Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions 12

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions • Dynamics & Autonomous UAV Control (single agent/enabling capabilities) – – – Trajectory/waypoint tracking Target tracking Collision and obstacle avoidance Integrated Guidance and Control (2) Vision-based control (6) • • • Active Contours Optic Flow Dynamic Feature Extraction 13

FY 03 MURI: Active-Vision Control Systems for Complex Adversarial 3 -D Environments • Air and ground vehicle tracking –Particle filtering + curve evolution to estimate the contour position and velocity for moving and deforming object –Optimal guidance policies for observability, including intelligent excitation –Integrated estimation/guidance with a composite adaptation approach • Obstacle/hazard avoidance –Layered active appearance models –Optical matting (separate Objective Develop methods that utilize 2 -D and 3 -D imagery to enable aerial vehicles to autonomously detect and prosecute targets in uncertain complex 3 -D adversarial environments -- without relying on highly accurate 3 -D models of the environment Participants • Georgia Tech: E. Johnson, UCLA, MIT, VT back/foreground) –Guidance and estimation for obstacle avoidance • Vision to replace traditional sensors –Vision-only flight control –Vision aided approach and landing –Vision-aided inertial navigation 14

PORTFOLIO OVERVIEW Technical Summary Introduction Cooperative Multi-agent Dynamics and Control AFOSR Overview Portfolio Overview Task Allocation (2) Future Direction/ Ideas – Path Planning (11) Conclusions – Tracking (3) – State Estimation (2) – Network Theory/Architecture (6) – Information Theory (2) – Mixed Initiative (2) Task Allocation Path Planning Target Tracking Decisions/ Computation Decentralized – Centralized • Management Summary • Technical Summary 15

FY 01 MURI Cooperative Control of Distributed Autonomous Vehicles in Adversarial Environments Introduction Goal: AFOSR Overview Deployment of Large Scale Networks of (semi) Autonomous Vehicles Portfolio Overview • Management Summary • Technical Summary Approach: Dimensions of Cooperative Control • Distributed control and computation • Vehicle flocking with obstacle avoidance • Optimal navigation in partially known environments • Adversarial interactions • Probabilistic differential games Future Direction/ Ideas • Mixed integer LP methods Conclusions Complex Collective Behavior from Simple Individual Behavior Sample Result Random rewiring of links with probability p increases performance of consensus algorithms and distributed filtering 1000 times • Uncertain evolution • Probability maps with moving opponents • Complexity management • Decomposition methods for hierarchical planning • Experimentation: • Case Study Simulations + Hybrid Hardware Realization Participants UCLA: J. Shamma, MIT, Caltech, Cornell 16

FY 02 MURI Cooperative Networked Control of Dynamical Peer-to-Peer Vehicle Systems Objective Control & Information Theory Computing & Verification Autonomous Vehicles Communications Scientific Approach • Scalable algorithms for verification of multivehicle systems • Languages for real-time networked vehicle interaction • Theory for information management in distributed feedback systems • Algorithms for allocations based on spatial geometry Participants • UIUC: G. Dullerud, MIT, Stanford • Establish theory, scalable algorithms and distributed protocols for achievable global performance in cooperative networked control. • Verify robustness to: uncertainty, malicious attacks, rapidly evolving mission objectives Accomplishments • Deployment Algorithms • Provable guarantees for coverage • New rigorous target servicing • Verification and validation • Switching for stochastic hybrid systems • Verification via learning and randomization • Control-oriented communication & information theory • Channel capacity theorem for control • Delay adaptive routing protocols 17

FY 07 MURI Behavior of Systems with Humans and Unmanned Vehicles Introduction AFOSR Overview • Waiting for competition results to be released. Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions 18

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Aerodynamic flow control and control of unsteady phenomena Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas • • Conclusions • • Lower order modeling (4) Control schemes (4) – Classical, optimal – Adaptive Sensor/Actuator placement (1) – Controllability/Observability issues Actuators – Synthetic jets, surface deflection, plasma 19

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Active waveform control (2) • Control of electromagnetic surface properties • Control of deformable mirrors and beam control Conclusions Initial Output Simulation Initial Output Experiment Final Output 20

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Dynamics and Modeling Portfolio Overview • • Management Summary • Technical Summary – – Future Direction/ Ideas Conclusions System Modeling (7) • • Wave dynamics for engines Atomic scale processes, Quantum control theory System Identification (1) Uncertainty Characterization (1) flutter Acoustics Structures thermoacoustics F(p, q) + a 2(q)pqq Wave Speed Combustion Fluid Dynamics Mistuning 21

PORTFOLIO OVERVIEW Technical Summary Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions General control theory • • Adaptive Control (5) Nonlinear Control (2) Hybrid Control (2) Other (4) 22

PORTFOLIO OVERVIEW Technical Summary Introduction Validation & Verification (3) AFOSR Overview Portfolio Overview Toolchain Model Certificates • Management Summary • Technical Summary • Example control design problem: Simulink/Stateflow Control Design Future Direction/ Ideas Conclusions + Simulink/ Stateflow Metamodels Component Model Platform Mapping Platform Topology • The nominal controller Knom is the LQR optimal feedback controller with double precision floating -point coefficients. • Admissible controllers C are controllers that yield an LQR cost that is, at most, 15% suboptimal. • The complexity measure Ф is the number of bits required to express K. • The best design, which is 14. 9% suboptimal, gives only 1. 5 bits/coefficients. = Target Code For RT System Configuration Files RT schedules Analysis Files Verification Models 23

FY 06 MURI: High Confidence Design for Distributed, Embedded Systems Objective Goal New feedback-based approaches to Develop new approaches to designing/developing distributed embedded systems that are designed systems to inherently promote high around V&V confidence, as opposed to design-then-test approaches as prescribed by the current V&V process Scientific Approach • Formal reasoning about distributed, dynamic feedback systems • Relationships between test coverage and system properties • Architectures to provide behavior guarantees of *online* V&V • V&V aware architectures • Multi-threaded control • Approximate V&V Figure 1 – Exponential Growth of Flight. Safety-Critical Systems Is Expected due Primarily to Autonomy Frameworks and Tools for High. Confidence Design of Adaptive, Distributed Embedded Control Systems Specification, Design and Verification of Distributed Embedded Systems 24

Future Directions/ Ideas Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas • Big problems that need work – – Conclusions Validation & Verification Mixed Initiative Cooperative Control Network & Information Theory in Controls Dynamics/Modeling/Uncertainty • • • – dynamic data dimensionality reduction techniques stochastic modeling of non-stationary dynamics classification of dynamical models of highdimension Future Directions Report (Spring 07) 25

Conclusions and Future Directions Introduction • AFOSR Overview Cohesive and well-connected program with national leadership, scientific innovations & technology transitions – Honors and accomplishments reflect research quality • • Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions – • 13 professional society fellows 2 NAE members 3 AFOSR star teams Robocup F 180 class world champions, 2003/2002/2000/1999 9 NSF career awards PECASE award winners in 2000, 2002 180+ reported refereed journal articles A priori consideration of practical application aids opportunistic technology transition Future directions in Dynamics & Control – Continue to pursue scientific advances in control for high risk, long range multidisciplinary and unconventional applications 26

Backup slides

Some Research Highlights Adaptive Flight Control in L-JDAM Introduction AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Some Research Highlights Future Direction/ Ideas • Reconfigurable Adaptive Flight Control with Limited Authority Actuation • Adaptive autopilot augmentation designs (tech transitions): • L-JDAM, (Laser guided MK-82 JDAM) – Demonstrated (in flight) fast adaptation to unknown aerodynamics – 2 successful flight tests at Eglin AFB Conclusions • November 2004: canned • January 2005: guided, stationary target – Summer 2006: successful flight tests, moving target Adaptive Flight Control Transitions 28

Some Research Highlights First-ever Vision Guided Autonomous Formation Flight Introduction Johnson, Georgia Tech, UCLA, MIT, VT AFOSR Overview Portfolio Overview • Management Summary • Technical Summary Some Research Highlights Future Direction/ Ideas Conclusions 30

Some Research Highlights Vision Guided Autonomous Tracking of Ground Vehicle Introduction Beard/Mc. Lain, Brigham Young Portfolio Overview • Management Summary • Technical Summary Trends in Dynamics & Control Some Research Highlights Future Direction/ Ideas Conclusions 31

Some Research Highlights Coordinated Autonomous Tracking/ Indoor flight lab Introduction How, MIT Portfolio Overview • Management Summary • Technical Summary Trends in Dynamics & Control Some Research Highlights Future Direction/ Ideas Conclusions 32

Discovery Challenge Thrusts (DCT) Introduction AFOSR Overview • • Systems and Networks • • • Radiant Energy Delivery and Materials Interactions • • Socio-Cultural Prediction Portfolio Overview • Management Summary • Technical Summary Future Direction/ Ideas Conclusions Integrated Sensors, Algorithmic Processors & Interpreters (I-ATR) Thermal Transport Phenomena and Scaling Laws Super-Configurable Multifunctional Structures Robust Decision Making Self-Reconfigurable Electronic/Photonic Materials & Devices Turbulence Control & Implications Space Situational Awareness Devices, Components, and Systems Prognosis 33

PORTFOLIO OVERVIEW Dynamics & Controls Introduction AFOSR Overview Jan Funding Timeline Portfolio Overview • Management Summary • Technical Summary White Papers/ Proposal Prep Note: AFOSR BAA is continuously open. Proposals can always be submitted at any time. However, in practice there is a timeline. Future Direction/ Ideas Conclusions Jun Practical Deadline for Proposals, 1 Jun External Reviews Oct Beginning of Fiscal Year, 1 Oct Dec Projected Grant Start Date, 1 Dec Funding Decisions 34