Dr. Michael Wolf is the supervisor of the Maritime & Multi-Agent Autonomy Group and a Principal member of the robotics technical staff. He serves as the principal investigator for several research projects in vehicle autonomy and multi-robot systems / swarms. These research efforts emphasize realistic applications of autonomy, using full-size vehicles, real-world communications and sensing, and maintainable and adaptable software. Dr. Wolfs technology contributions span perception, motion planning, multi-vehicle coordination, autonomy software architectures, marine robotics, and wearable human–robot interfaces, generally emphasizing probabilistic methods for estimation, tracking, and autonomous control.
* PhD in Robotics / Mechanical Engineering, California Institute of Technology, 2008
Thesis in robotics: "Target Tracking Using Clustered Measurements, with Applications to Autonomous Brain-Machine Interfaces"
Minor in Control and Dynamical Systems
* MS in Mechanical Engineering, California Institute of Technology, 2005
* BS in Mechanical Engineering, Stanford University, 1997
* Group Supervisor, JPL Maritime & Multi-Agent Autonomy, Robotics Section, 2018-Present
* Robotics Technologist, JPL Robotics Section, 2008-2018.
* Adjunct Associate Professor, Art Center College of Design, 2007-Present
* Six years experience in technical consulting, product management, and business development for enterprise software companies.
JPL Projects with a significant role:
* USV (Unmanned Surface Vehicle) Swarm—PI (2014–2020) and Task Manager (2014–2017)
* Sea Mob Autonomous Swarm Development and Demonstration—PI and Task Manager (2015–2019)
* Autonomy for ASW Continuous Trail Unmanned Vehicle (ACTUV)—PI and Task Manager (2013–2017)
* Autonomous Perception and Hazard Avoidance—PI (2015)
* Advanced Sensing and Behaviors for the ACTUV—PI and Task Manager (2013–2014)
* JPL Autonomy for USSV for LCS Missions—Task Manager (2008–2014)
* ACTUV Multi-Objective Optimization and Control—PI and Task Manager (2010–2012)
Increased autonomy capabilities of uncrewed boats across these tasks, starting with motion planning and hazard avoidance, then increasing mission and behavior autonomy, then full mission demonstrations with cooperative teams of vehicles. Invented new autonomy algorithms. Reinvented the CARACaS 2 autonomy software suite and matured to TRL 7.
* DARPA Subterranean Robotics Challenge—Co-I and Deputy Task Manager (2018–2021)
Deploy a heterogeneous team of robots into underground environments (mines, subways, and caves) to autonomously explore and map the volume to very high accuracy, with challenging communications and time limits.
* Distributed event-driven task determination and allocation in robot teams—PI (2018–2020)
Solve key decision-making steps in multi-robot, event-driven missions, where tasks are often not known a priori and need to be determined and allocated based on information collected in situ by the robotic agents.
* Mechanically Uncoupled Stereo EO/IR (MUSE)—PI (2015–2017)
* Advanced Contact Detection and Analysis for USV Autonomy—PI and Task Manager (2013–2015)
Game-changing new camera-based perception technologies, including self-calibrating, real-time stereo processing between independently mounted cameras, enabling very long-range object detection and ranging, and machine learning for classification of objects in the environment.
* Wearable Interface for Natural Control and Tele-Operation of Robots—Co-I (2014–2016)
* Bioelectric Signal Arrays for Reliable Prosthetics Interfaces—Co-I (2012–2014)
Novel forearm “BioSleeve” that captures hand and arm gestures via electromyography (EMG) and inertial measurements; led machine learning for bio-signal classification and sliding-autonomy interaction for robot and prosthesis control.
* Lunar Surface Operations Testbed—Algorithm Lead, Sample Verification (2010–2011)
JPL testbed for lunar sample return mission; invented methods to estimate the mass and distribution of regolith contained in a sample collection cannister, based on noisy readings from a custom 7-channel capacitive sensor.
* SRTD: Exploitation of Aerial Imagery—Algorithm Lead, Pedestrian Tracking (2011–2012)
* Angel Fire—Software Module Lead, Landmark Tracking (2010–2011)
Detection, tracking, and classification of ground objects from airborne cameras; invented new integrated detection/ tracking algorithms for locating pedestrians in poorly stabilized, noisy imagery and provided landmark tracking to support real-time geo-registration and high resolution 3D terrain generation.
* Titan Aerobot—Algorithm Lead, Probabilistic Motion Planning (2008–2010)
Trade study for system design of potential Titan Montgolfiere aerobot; developed new probabilistic methods for 3D wind- assisted motion planning to address the uncertain atmospheric forecast models.
* Robot Autonomy decision making, motion planning, multi-agent coordination, etc.
* Perception probabilistic estimation and tracking
* Autonomy Software Architectures