Token based path allocation approach for mobile robots in sorting centre for achieving high throughput

• Main Author: Khoo, Zong Zheng
• Format: Final Year Project / Dissertation / Thesis
• Published: 2024
• Subjects: T Technology (General); TA Engineering (General). Civil engineering (General); TD Environmental technology. Sanitary engineering
• Online Access: http://eprints.utar.edu.my/6644/1/fyp_CS_2024_KZZ.pdf; http://eprints.utar.edu.my/6644/
• Institution: Universiti Tunku Abdul Rahman

As the trend of online shopping and e-commerce rises rapidly after the COVID-19 pandemic, warehouses and sorting centres are facing challenges of managing an enormous number of parcels on a daily basis. To cope with these demands, changing their warehouse to automation are one of the solutions that these companies apply. Autonomous Mobile Robots (AMR) or Automated Guided Vehicle (AGV) is applied to these automated warehouses to sort and deliver these parcels to the desired destination. Normally, these robots are deployed in large quantities. Thus, it is crucial to guaranteed that these robots avoid collisions and deadlocks among each other while executing tasks. There are some previous approaches that provides different sets of assumptions on the road systems, planning algorithms and layouts. However, some of these assumptions does not obey real-life phenomena such as inertia. Thus, it may be hard to apply this proposed solution in the use of real-life. Also, there are papers that proposes algorithms that are able to solve deadlocks and collision among these robots but not able to solve the heavy traffic problem while passing through an intersection grid cell. Heavy traffic will affect the throughput of completing the tasks. Thus, these problems seem to be the reasons that may be not that applicable and inefficient to apply in the real world. From here, we will first review and evaluates some of the approaches that have been done by previous studies. These papers will be reviewed to find some inspiration and discover some potential new approaches that might be able to handle multiple robots at the same time while decreasing the traffic as much as possible. This project proposes an improved path-finding algorithm that implements token-based path finding while a robot wants to access a grid cell. Besides that, the projects also propose to form and break clusters when these robots are moving in a same direction to decrease the queueing time and increases the throughput. Also, a layout is proposed that implements ring road and/or central road concept which allow these robots to travel using these roads without passing the main road that links to the pick-up and drop-off areas.