Name
Abstract | ||
|---|---|---|
With the rising globalisation and specialisation of today's manufacturing industries, the significance of telepresence systems is gaining increased attention. In such applications, robotic devices called teleoperators are controlled from remote locations by human operators, who are supplied with all relevant input devices and feedback information to achieve the task at hand. Considering the distributed nature of telepresence environments, the presence of network delays is inevitable. Between any two communication parties, delays in data transmission will exist. Depending on the specific scenario, this delay is often much larger than desired. A way in which data receiving clients can attempt to minimise the effects of delays, is to predict what the future data points will be. This research investigates the use of the support vector regression machine learning algorithm to predict future position commands sent from a 6 degree of freedom Phantom haptic input device to a remote microassembly teleoperator. In particular, the impact of network delay on prediction accuracy is analysed, and a method of 'partial prediction' is shown to improve the synchronisation between communicating telepresence partners in high network delay scenarios. |
Year | Venue | Keywords |
|---|---|---|
2003 | HIS | telepresence application,support vector regression,haptic data |
DocType | Volume | ISSN |
Conference | 104 | 0922-6389 |
ISBN | Citations | PageRank |
1-58603-394-8 | 3 | 0.59 |
References | Authors | |
2 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Stella M. Clarke | 1 | 51 | 4.84 |
| Michael F. Zaeh | 2 | 109 | 20.34 |
| Jan H. Griebsch | 3 | 3 | 0.59 |