Autonomous Warehouse Logistics with Multiple Mobile Manipulators

Project: Research

View graph of relations


Warehouse logistics has been the bottleneck in many industrial applications. For instance, in e-commerce, logistics creates problems such as slow and/or wrong deliveries, lost packages, damaged goods, and incorrect packing. In manufacturing, logistics tasks such as transportation of parts between workstations or multiple/single part feeding [4] are also the components that restrict the robustness and efficiency of the production line. Traditional automation/robotics techniques have been leveraged to extend capacities and capabilities of the industrial logistics, such as the automated storage and retrieval system (ASRS) for efficient e-commerce delivery, and the industrial robots for performing dumb, dangerous, dull and dirty logistics tasks. Despite the high efficiency provided by these existing systems, they are dedicated and/or fixed, and thus are limited in flexibility when they come to dealing with objects or parts of various size, shape, weight, volume, and mechanical properties. In addition, these systems may not offer adequate scalability to adapt to growth and cope with increased seasonal demands, or deal with facility breakdowns and carry out technical maintenance. The autonomous mobile manipulator – a robot manipulator mounted upon a mobile platform, extended by vision, tooling, and AI-based decision making systems – promises to strike a balance between efficiency, scalability, and flexibility, by combining the locomotion capabilities with manipulation abilities. The mobility extends the workspace of the robot manipulator, which increments the robot’s operational capability and flexibility. Compared to traditional automation system or industrial robots, it is easier for mobile manipulators to adapt to changing environments and perform a wide variety of logistics tasks. In addition, while deploying multiple mobile manipulators, the factory/warehouse environments do not have to be altered as in the case of automated guided vehicles (AGV) – with Kiva systems currently used in Amazon warehouses as an example – where permanent cable layouts, computerized barcode stickers on the floor or bounce lasers off reflectors are necessary for robot localization and navigation. Therefore, autonomous mobile manipulators have the potential to be “ready to go right out-of-the-box” and will reduce the deployment time and cost of the automated warehouse logistics system.


Project number9048077
Grant typeECS
Effective start/end date1/01/172/01/19

    Research areas

  • automated warehouse , task planning , multi-agent navigation , motion planning , grasping and manipulation