Degree Name



School of Electrical, Computer and Telecommunications Engineering - Faculty of Informatics


An empirical research on developing a new paradigm for programming a robotics manipulator to perform complex constrained motion tasks is carried out in this thesis. The teaching of the manipulation skills to the machine commences by demonstrating those skills in a haptic-rendered virtual environment. This is in contrast to conventional approach in which a robotics manipulator is programmed to perform a particular task. A manipulation skill consists of a number of basic skills that, when sequenced and integrated, can perform a desired manipulation task. By manipulation means the ability to transfer, physically transform or mate a part with another part. Haptic-rendering augments the effectiveness of computer simulation by providing force feedback for the user. This increases the quality of human - computer interaction and provides an attractive augmentation to visual display and significantly enhances the level of immersion in a virtual environment. The study is conducted based on the peg-in-hole application as it concisely represents a constrained motion-force-sensitive manufacturing task with all the attendant issues of jamming, tight clearances, and the need for quick assembly times, reliability, etc. The state recognition approach is used to identify and classify the skills acquired from the virtual environment. A human operator demonstrates both good and bad examples of the desired behaviour in the haptic virtual environment. Position and contact force and torque ii data, as well as orientation generated in the virtual environment, combined with a priori knowledge about the task, are used to identify and learn the skills in the newly demonstrated tasks and then to reproduce them in the robotics system. The robot evaluates the controller’s performance and thus learns the best way to produce that behaviour. The data obtained from the virtual environment is classified into different cluster sets using the Hidden Markov Model (HMM), Fuzzy Gustafson–Kessel (FGK) and Competitive Agglomeration (CA) respectively. Each cluster represents a contact state between the peg and the hole. The clusters in the optimum cluster set are tuned using a Locally Weighted Regression (LWR) algorithm to produce prediction models for robot trajectory performing the physical assembly based on the force/position information received from the rig. The significance of the work is highlighted. The approach developed and the outcomes achieved are reported. The development of the haptic-rendered virtual peg-in-hole model and structure of the physical experimental rig are described. The approach is validated though experimental work results are critically evaluated. Keywords: Haptic, PHANToM, ReachIn, Virtual Reality, Peg-in-hole, Skill acquisition.

02Whole.pdf (3472 kB)