Skip to main content

PhD candidate Mithun Goutham showcases advancements in path planning at the IROS 2023 Conference

Posted: 

I recently had the privilege to present one of the many research outcomes of a Ford-OSU Alliance project at the 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2023) which was held in Detroit, Michigan. The research project, which focused on exploring flexible manufacturing systems (FMS), was led at The Ohio State University by my advisor, Professor Stephanie Stockar.

Enhancing industrial resilience in the face of unpredictable challenges like the COVID-19 pandemic is increasingly of interest to companies

Meghna Menon and Goutham at the conference

like the Ford Motor Company as they navigate toward Industry 4.0, the next industrial revolution. The flexibility offered by FMS is seen as a crucial element to this endeavor, and one of its most important enablers is the transition of material handling operations away from traditional assembly lines to fleets of autonomous mobile robots (AMRs).

Efficiency was a crucial factor in our optimization framework from the outset, leading us to uncover an oversight in the conventional approach of routing AMRs. Common practice relied on point-to-point traversal costs in path planning algorithms, but this failed to account for additional expenses incurred during turns, stops at intersections, and merges into regions of high congestion.

By employing an efficient representation of the factory environment using a dual-graph, and by simultaneously considering all waypoints that an AMR would encounter on its material handling mission, we successfully demonstrated a tractable method to reduce the true path costs. We documented our findings in a manuscript titled Optimal Path Planning Through a Sequence of Waypoints which was accepted to the IEEE RA-L journal, and then recommended to the IROS 2023, a flagship conference of the IEEE Robotics and Automation Society.

The conference was an inspiring blend of innovative research ideas and the latest in robotics, showcased by industries and laboratories worldwide. The sheer variety of the robotic exhibits was honestly overwhelming and showed the vast extent of tangible innovation currently underway. It was also incredible to attend workshops and have conversations with experts in the field of swarm robotics and fleet control, especially since these are areas I am actively working on.

My presentation was well-received, and the poster session proved to be particularly enlightening. I found people keen to know whether off-the-shelf path planning algorithms could be applied directly, and some inquired whether I knew of a heuristic for the A* algorithm if applied in this context. Attendees were enthusiastic to discuss how this could be extended to their fields, such as aerial applications, ship navigation and racing.

Thanks to Meghna Menon, the Ford Principal Investigator, I had the opportunity to tour the Ford Manufacturing Technology Development Center where research is conducted on cutting-edge manufacturing innovations such as automated vehicle marshaling. She was recently invited to be a seminar speaker at CAR, and her insightful presentation titled “Leveraging Multi-Robot Systems for the Factory of Tomorrow” can be viewed here. I am very grateful for the perspectives I have gained from our collaboration, which merged theory with practical applications and enriched my understanding of the field.

Written by Mithun Goutham, PhD candidate, Mechanical and Aerospace Engineering