Project: Improving robotic manipulation through intelligent multimodal perception
Robotic systems have challenges to grasp and manipulate objects. The grasping process could be described in three main stages: Detection, planning and Execution. The foundation to complete a task is to have Generalizing methods for the new objects, environment and tasks which could be quite difficult due to uncertainty and lack of appropriate perception. The foundation to complete a task is to have Generalizing methods for the new objects which could be quite difficult. This research will seek a novel method of manipulation by combining tactile sensing methods, vision data and machine learning techniques. The first step will be to collect the data and analyze the data in the next step to finally introduce a novel and cost-effective method. This work will be in combining tactile and vision to improve complex tasks. By using the tactile and vision data, precision and flexibility, as well as repeatability, will improve. The main problem to be addressed during this research will be how to leverage tactile and vision to improve robotic manipulation. Uncertainties and the working environment in which the robotic system will be functioning should be taken into consideration.
Supervisor(s):
