Enabling methodologies for optimal coverage by multiple autonomous industrial robots

Publication Type:
Thesis
Issue Date:
2018
Metrics:
Full metadata record
Files in This Item:
Filename Description Size
01front.pdf169.98 kB
Adobe PDF
02whole.pdf12.84 MB
Adobe PDF
Unlike traditional industrial robots which are purpose-built for a particular repetitive application, Autonomous Industrial Robots (AIRs) are adaptable to new operating conditions or environments. An AIR is an industrial robot, with or without a mobile platform, that has the intelligence needed to operate autonomously in a complex and unstructured environment. This intelligence includes aspects such as self-awareness, environmental awareness, and collision avoidance. In this thesis, research is focused on developing methodologies that enable multiple AIRs to perform complete coverage tasks on objects that can have complex geometric shapes while aiming to achieve optimal team objectives. For the AIRs to achieve optimal complete coverage for tasks such as grit-blasting and spray painting several problems need to be addressed. One problem is to partition and allocate the surface areas that multiple AIRs can reach. Another problem is to find a set of appropriate base placements for each AIR and to determine the visiting sequence of the base placements such that complete coverage is obtained. Uncertainties in base placements, due to sensing and localization errors, need to be accounted for if necessary. Coverage path planning, i.e. generating the AIRs’ end-effector path, is another problem that needs to be addressed. Coverage path planning needs to be adaptable with respect to dynamic obstacles and unexpected changes. In solving these problems, it is vital for the AIRs to optimize the team's objectives while accounting for relevant constraints. This research develops new methodologies to address the above problems, including (1) a Voronoi partitioning based approach for simultaneous area partitioning and allocation utilizing Voronoi partitioning and multi-objective optimization; (2) optimization-based methods for multi-AIR base placements with uncertainties; and (3) a prey-predator behaviour-based algorithm for adaptive and efficient real-time coverage path planning, which accounts for stationary or dynamic obstacles and unexpected changes in the coverage area. Real-world and simulated experiments have been carried out to verify the proposed methodologies. Various comparative studies are presented against existing methods. The results show that the proposed methodologies enable effective and efficient complete coverage by the AIRs.
Please use this identifier to cite or link to this item: