A robotic workcell is a defined area of space through which the robot can move and perform the required task. Up to now, the industry still faces three major challenges on setting up a robotic workcell for automation tasks. The first challenge is the rudimentary effort for workcell layout planning, i.e., positioning the robot and the workpiece according to the task: the robot trajectories, workspace, and environment constraints.
Currently such layout planning is carried out mainly based on experience and trial-and-error. When different types of workpieces or robot configurations are introduced, the same manual effort is repeated again and again. Although robotizing the manufacturing processes save time and manpower cost and improve the productivity of the overall work, the effort in setting up the workcell remains no improvement. This problem will be even more critical if the robot is deployed in an unstructured field environment.
The second challenge is the variation and compensation of actual workpiece geometry, weight, and location on the fixture right before processing. The current practice in a high volume manufacturing setting is to simply reject such workpieces. For low volume and high mix manufacturing, there is no readily available system that can automatically detect these variations, re-adjust the workpiece, and re-program the robot once they are set up in the workcell. The third challenge is the lack of an agile robotic workcell that can be quickly configured, set up, scaled, and operated on very large workpieces. For example, field welding practiced in several key Singapore industries is a tiring, laborious and foreign-worker intensive task. It is greatly desired that adaptive robot welding workcell can be incorporated into industries like shipyards and offshore oil and gas equipment makers to replace this bottleneck manual-welding task. Existing robots work well in the structured indoor environment with controlled ambient conditions but not the outdoor unstructured environment. This calls for new and innovative robotic solutions at the workcell level.
The objective of WP 5 - adaptation - is to develop software and hardware platform technology in robotic workcell to address the quick layout planning and set-up of workpieces and the robots, robust compensation for part variations and operational uncertainty at workcell level, and the ability to adapt to different working environments for very large workpieces in the field. Specific aims are:
To achieve automated task-oriented workcell layout planning under both structured and unstructured environments to facilitate fast workflow in setting up the robotic manufacturing system.
To enable online workpiece registration and localization based on actual part variations at workcell level for robust operation with minimum rejection.
To achieve fully automated and integrated in-situ robot calibration for field robotics workcell.
To develop adaptive mobile workcell prototypes that can be configured to move along supporting infrastructure and reach difficult access areas, for field operations like welding, inspection and painting on very large workpieces.