Aeronautics & Astronautics Autonomous Flight Systems Laboratory All slides and material copyright of University of Washington Autonomous Flight Systems Laboratory

Aeronautics & Astronautics Autonomous Flight Systems Laboratory Mission Planning and Execution for Unmanned Surface Vehicles in Compliance with the Marine Rules of the Road A Master’s Thesis by Jim Colito AERB 117 (206)

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington3 Outline Problem statement Motivation for USV technology development Current USV technology Short description of path planner Marine Rules of the Road – description and implementation Results

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington4 Problem Statement USVs are not as technically mature as UAVs Primarily human environment is highly unpredictable → USV must be flexible and agile Need to prevent injury and loss of property from USV introduction into waterways Current technology requires constant decision making by operators A primary technical challenge is obstacle avoidance in a diverse, highly dynamic environment Two regimes of obstacle avoidance: near-field (< yards) and far- field Include “Marine Rules of the Road” in path planning system to reduce liability and address an aspect of far-field obstacle avoidance behavior.

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington5 Why USVs? Why AFSL??? Potential benefits to Intelligence Surveillance and Reconnaissance (ISR) missions  Interdiction, IED identification, Surveying, Research, Search and Rescue, Commercial Fishing, etc… Recent ONR briefing identified the following benefits Minimize risk to personnel in high-risk littoral missions Low cost Not power limited Not limited by human factors Ability to communicate with UUVs and UAVs Attacks on USS Cole (2000), oil tanker Limburg (2002), Phillipine Superferry 14 (2004) and Khor Al Amaya oil terminal(2004) Unmanned technology is well-suited for Dull, Dirty, and Dangerous jobs

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington6 Outline Problem statement Motivation for USV technology development Current USV technology Short description of path planner Marine Rules of the Road – description and implementation Results

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington7 Typical USV Technology In the past primarily used as radio controlled target training devices Typically personal watercraft size or smaller Expensive Lack comprehensive mission management and obstacle avoidance OWLROBOSKI

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington8 USV Technology Continued PROTECTOR SSC - SAN DIEGO SPARTAN SCOUT

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington9 Platform SEAFOX Platform  Durable  Easy to transport  Quick set-up  Relatively cheap

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington10 Outline Problem statement Motivation for USV technology development Current USV technology Short description of path planner Marine Rules of the Road – description and implementation Results

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington11 Overall Mission Scenario

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington12 Commit to a trajectory Continually plan next path Commit to the beginning of the best path at the start (spawn) point DYNAMIC REPLANNING Path Generation The path planner uses an evolutionary process to plan paths. The steps in the evolution are: 1.) Evaluates Fitness of paths 2.) Selects the best paths 3.) Produce offspring from the best paths to be mutated and evaluated

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington13 Outline Problem statement Motivation for USV technology development Current USV technology Short description of path planner Marine Rules of the Road – description and implementation Results

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington14 Rules of The Road Overview Using the Marine Rules of the Road as defined by the COLREGS increases safety and should reduce liability. Most rules defined by COLREGS relate to lighting, signals, definitions, etc… Rules that define Head-On, Crossing, and Overtaking Collisions are important. Other Rules that need to be considered are Identifying a collision When to take action to avoid collision Right-of-way → assume our vehicle cedes right-of-way!

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington15 COLREGS Interpretation Rule 7a (risk of collision) – Every vessel shall use all available means appropriate to the prevailing circumstances and conditions to determine if risk of collision exists. If there is any doubt such risk shall be deemed to exist. What is doubt? Rule 8a (Action to Avoid Collision) – Any action taken to avoid collision shall, if the circumstances of the case admit, be positive, made in ample time and with due regard to the observance of good seamanship. Rule 8b (Action to Avoid Collision) – Any alteration of couse and/or speed to avoid collision shall if the circumsances of the case admit, be large enough to be readily apparent to another vessel observing visually or by radar; a succession of small alterations of course and/or speed should be avoided. Ample time? Observance of good seamanship? Large enough? Need to keep these rules in mind when designing a cost function!!!

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington16 Collision Scenarios Vehicle-to-vehicle there are three basic basic collision scenario’s Head-on collision Crossing collision Overtaking collision Crossing Head-On Obstacle Head-On 120 º 15 º Overtaking

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington17 Head-On Collision Scenario Safety Radius Noncompliant Change of Course Safety Radius COLREG Compliant Change of Course Definition: Rule 14a – When two power-driven vessels are meeting on reciprocal or nearly reciprocal courses so as to involve risk of collision each shall alter her course to starboard so that each shall pass on the port side of the other. Rule 14b – Such a situation shall be deemed to exist when a vessel sees the other ahead or nearly ahead and by night she could see the masthead lights of the other in a line or nearly in a line and/or both sidelights and by day she observes the corresponding aspect of the other vessel. Rules written to require interpretation by operator Involve identification of lighting We ought to pass with port sides facing each other.

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington18 Crossing Collision Scenario Safety Radius Undesirable Crossing Behavior Safety Radius Desired Crossing Behavior Definition: Rule 15 – When two power-driven vessels are crossing so as to involve risk of collision, the vessel which has the other on her own starboard side shall keep out of the way and shall, if the circumstances of the case admit, avoid crossing ahead of the other vessel. Desirable to pass behind other vehicles. We assume our vessel is always the vessel passing behind!

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington19 Implementation of Rules First, determine the closest moving object in the rule range Active Obstacle Inactive Obstacle 120 º Rule Range

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington20 Collision Cone *Proved in A. Chakravarthy and D. Ghose “Obstacle Avoidance in a Dynamic Environment: A Collision Cone Approach” R VoVo r VTVT φ α β Check if any of these 7 points are in collision cone Candidate Path #1 Candidate Path #3 Candidate Path #2 (Best Path) Committed Trajectory Collision Cone Evaluation Points Spawn Point

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington21 What’s the Navigation Rule? Head -On Crossing Head-On Obstacle Head-On 120 º 15 º

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington22 Discretization of Head-On scenario Candidate Trajectory 1 Candidate Trajectory 2

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington23 Evaluation of Head-On α 11 α 12 α 22 Candidate Trajectory 1 Candidate Trajectory 2 Construct Port Vector Calculate angle between Port Vector and connecting line

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington24 Evaluation of Crossing Scenario Candidate Trajectory 1 Candidate Trajectory 2 α 11 α 12 O TOTO Same cost function as head-on scenario! Construct Stern Vector instead of Port Vector!

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington25 Outline Problem statement Motivation for USV technology development Current USV technology Short description of path planner Marine Rules of the Road – description and implementation Results

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington26 Results Head-on collision scenario constructed to encourage COLREG noncompliant passing Path planner originally plans to pass on starboard side but changes to port

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington27 Results Cont… Crossing collision scenario constructed to encourage COLREG noncompliant crossing Path planner originally plans to pass in front of obstacle

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington28 Results Cont… Include terrain data… Task: Visit two targets while avoiding boat traffic in a COLREG compliant manner.

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington29 Conclusions

Aeronautics & Astronautics Autonomous Flight Systems Laboratory University of Washington30 Thanks To… Dr. Rolf Rysdyk Dr. Juris Vagners Dr. Anawat Pongpunwattana Northwind Marine Inc. Washington Technical center Autonomous Flight Systems Laboratory Guggenheim 109 (206) I’d like to thank the following people/organizations for giving me the opportunity to work on this project: