Perceptual haptic data communication for telepresence and teleaction

Telepresence and teleaction (TPTA) systems enable a human operator to execute complex tasks within an inaccessible, dangerous, and/or scaled environment. Their goal is to provide a local and realistic immersion of the remote environment. In this context, the communication of haptic signals (velocity and force) plays an important role. As the exchange of haptic information over the communication channel closes a global control loop between the human and the remote environment, the design and the development of methods for perceptual haptic communication must be carefully adapted to expected communication latency and network performance. This dissertation presents a broad spectrum of methods for the perceptual communication of haptic signals. Challenging in this context is the design of an accurate psychophysical model of human haptic perception. It enables the mapping of neural excitements to a perception sensation in the human brain which can be used to reveal absolute and relative perception thresholds. This allows for controlling the intensity of compression artefacts in order to keep them below human haptic perception thresholds. Accordingly, the integration of additional findings frompsychophysics into the perceptualmodel promises improved adaptation to the haptic discrimination thresholds which can be exploited in the context of perceptual haptic data compression. In addition, the application of environment models in haptic communication is proposed. Specifically, the bidirectional nature of haptic communication in TPTA systems enables the estimation of the geometry and impedance characteristics of the remote surface in contact. This enables the decoupling of the global haptic control loop which allows for improved transparency and compression performance. In addition to real-time communication of haptic signals, the development of offline compression methods is also of great interest. To this end, a perceptual haptic offline compression scheme is proposed which exploits knowledge about pending perception bounds resulting in an additional entropy minimization of the compressed data. Furthermore, to enable a theoretical analysis of the proposed methods for perceptual haptic communication, a mathematical framework based on stochastic models of haptic signals is discussed. Additionally, the impact of communication uncertainties such as communication latency and packet loss is addressed. Here, a modified control scheme for time-delayed haptic communication is presented which ensures stable operation and an efficient utilization of the haptic communication channel. Furthermore, a novel low-delay and perceptually-motivated haptic communication scheme is proposed, which is robust to data loss on the haptic channel. In order to manage and control TPTA sessions, as well as to facilitate the integration of the developed perceptual communication methods into the haptic communication system, a flexible communication framework is proposed which is based on standard Internet communication protocols. All developed methods and architectures are validated in experimental user studies.