Research Publications
Publications
-
Constrained Nonlinear Kaczmarz Projection on Intersections of Manifolds for Coordinated Multi-Robot Mobile Manipulation
Akshaya Agrawal, Parker Mayer, Zachary Kingston, and Geoffrey A. Hollinger
Accepted for 2025 IEEE International Conference on Robotics and Automation (ICRA), Atlanta, USA, 2025. (To Appear)
Cooperative manipulation tasks impose various structure-, task-, and robot-specific constraints on mobile manipulators. However, current methods struggle to model and solve these myriad constraints simultaneously. We propose a twofold solution: first, we model constraints as a family of manifolds amenable to simultaneous solving. Second, we introduce the constrained nonlinear Kaczmarz (cNKZ) projection technique to produce constraint-satisfying solutions. Experiments show that cNKZ dramatically outperforms baseline approaches, which cannot find solutions at all. We integrate cNKZ with a sampling-based motion planning algorithm to generate complex, coordinated motions for 3 to 6 mobile manipulators (18–36 DoF), with cNKZ solving up to 80 nonlinear constraints simultaneously and achieving up to a 92% success rate in cluttered environments. We also demonstrate our approach on hardware using three Turtlebot3 Waffle Pi robots with OpenMANIPULATOR-X arms.@inproceedings{agrawal2024cnkz, title = {Constrained Nonlinear {Kaczmarz} Projection on Intersections of Manifolds for Coordinated Multi-Robot Mobile Manipulation}, author = {Agrawal, Akshaya and Mayer, Parker and Kingston, Zachary and Hollinger, Geoffrey A.}, year = {2025}, booktitle = {IEEE International Conference on Robotics and Automation}, eprint = {2410.21630}, archiveprefix = {arXiv}, primaryclass = {cs.RO}, note = {To Appear}, } -
Underwater Multi-Robot Simulation and Motion Planning in Angler
Akshaya Agrawal, Evan Palmer, Zachary Kingston, and Geoffrey A. Hollinger
Accepted for 2025 OCEANS, Brest, France, 2025 (To Appear)
Deploying multi-robot systems in underwater environments is expensive and lengthy; testing algorithms and software in simulation improves development by decoupling software and hardware. However, this requires a simulation framework that closely resembles the real-world. Angler is an open-source framework that simulates low-level communication protocols for an onboard autopilot, such as ArduSub, providing a framework that is close to reality, but unfortunately lacking support for simulating multiple robots. We present an extension to emph{Angler} that supports multi-robot simulation and motion planning. Our extension has a modular architecture that creates non-conflicting communication channels between Gazebo, ArduSub Software-in-the-Loop (SITL), and MAVROS to operate multiple robots simultaneously in the same environment. Our multi-robot motion planning module interfaces with cascaded controllers via a exttt{JointTrajectory} controller in ROS~2. We also provide an integration with the Open Motion Planning Library (OMPL), a collision avoidance module, and tools for procedural environment generation. Our work enables the development and benchmarking of underwater multi-robot motion planning in dynamic environments. -
Task and Motion Planning for Collective Robot Construction
Akshaya Agrawal, Dongsik Chang and Geoffrey A. Hollinger
In ICRA 2022 Workshop on Collective Robotic Construction, May 27, 2022, Philadelphia, PA.
To improve construction efficiency and enable structure fabrication in hazardous terrestrial and underwater environments, we envision a team of autonomous mobile manipulators (AMMs) capable of assembling and re-arranging construction materials. In order to accomplish this goal, a large number of overlapping tasks and geometric constraints must be imposed on the AMM end-effectors to allow for cooperative manipulation. We propose a cooperative path planning algorithm: Intersection of Manifolds RRT∗ (IMaRRT∗), which enables complex cooperative manipulation by exploring an intersection of manifolds and solving for a large system of nonlinear constraints. We also discuss the existing challenges for determining the task order to assemble the structure, executing the planned path due to AMM and object dynamics, and incorporating environmental uncertainties. -
Exploring behavioral anthropomorphism with robots in virtual reality
Chinmay P Wadgaonkar, Johannes Freischuetz, Akshaya Agrawal, and Heather Knight
4th International Workshop on Virtual, Augmented, and Mixed Reality for HRI
Virtual reality (VR) and social robotics have mutual benefits. VR offers an instrumented and manipulable environment in which robots and people can virtually interact as well as tools for visual manipulations of robot materiality and color. VR also has a wealth of knowledge about how multimodal communications like motion, proxemics, and touch can inform interaction. Submersing social robots in VR provides an opportunity for physically-grounded interaction that leverages behavioral anthropomorphism. This work attempts to intersect these previously disparate areas, eliciting participant storytelling about the simplest possible anthropomorphizable robot: a robot that approaches and then bumps into you. In the study, 16 participants experience twelve manifestations of virtual/physical robots that approach and collide into them. The moment of collision provides an opportunity for expressive interpretation that offers a first glimpse into future potentials for physically embodied companion characters in virtual reality.@inproceedings{wadgaonkar2021exploring, author = {Wadgaonkar, Chinmay and Freischuetz, Johannes and Agrawal, Akshaya and Knight, Heather}, year = {2021}, month = {03}, pages = {}, title = {Exploring Behavioral Anthropomorphism With Robots in Virtual Reality} } -
Developing a bio-inspired pole climbing robot
M. Das, A. Agrawal, A. Sonone, R. Gupta, D. Upadhyay, Y.V.D. Rao, and A. Javed
2016 International Conference on Robotics: Current Trends and Future Challenges, 2016.
This paper proposes a pole climbing robot that has an ability to climb pipes. The novelty of this design is that it uses no motors to climb on the pole. Till now many robots have been fabricated with the ability to climb pipes but most of them solely depend on DC motors. The use of DC motor induces the risk of loosening the grip of the robot in case of power failure that may lead to disastrous situations if robot is working on high altitudes. This could be easily avoidable by the use of electro-pneumatics and by using self-locking circuits. We used two pairs of pneumatic cylinders as linear actuators for gripping the pole i.e. one pair for one claw. Besides this two more cylinders are used for climbing purpose. Unlike other mechanisms, installation is extremely easy. It is like controlling a radio controlled car. This mechanism can be easily modified for different payload capacities by simply using the appropriate diameter of the pneumatic cylinders. The robot can be controlled in tele-operated mode as well as autonomous mode which is a closed loop system. It has a high accuracy with 4mm error per cycle that has a cycle time period of 5 seconds. This robot can be very useful in oil industries for inspection of pipes and power poles in electric industries. It can serve of great help in cases of emergency situation caused during fire accidents.@inproceedings{das2016developing, title={Developing a bioinspired pole climbing robot}, author={Das, Moloy and Agrawal, Akshaya and Sonone, Abhinav and Gupta, Rishabh and Upadhyay, Deepak and Rao, YVD and Javed, Arshad}, booktitle={International Conference on Robotics: Current Trends and Future Challenges (RCTFC)}, year={2016}, pages={1-6}, doi={10.1109/RCTFC.2016.7893400}}