Cloud-based Networked Visual Servo Control

Haiyan Wu, Lei Lu, Chih-Chung Chen, Sandra Hirche, Kolja Kühnlenz

    Research output: Contribution to journalJournal articleResearchpeer-review

    1000 Downloads (Pure)


    The performance of vision-based control systems, in particular of highly dynamic vision-based motion control systems, is often limited by the low sampling rate of the visual feedback caused by the long image processing time. In order to overcome this problem, the networked visual servo control, which integrates networked computational resources for cloud image processing, is considered in this article. The main contributions of this article are i) a real-time transport protocol for transmitting large volume image data on a cloud computing platform, which enables high sampling rate visual feedback, ii) a stabilizing control law for the networked visual servo control system with time-varying feedback time delay, and iii) a sending rate scheduling strategy aiming at reducing the communication network load. The performance of the networked visual servo control system with sending rate scheduling is validated in an object tracking scenario on a 14 degree-of-freedom dual-arm robot. Experimental results show the superior performance of our approach. In particular the communication network load is substantially reduced by means of the scheduling strategy without performance degradation.
    Original languageEnglish
    JournalI E E E Transactions on Industrial Electronics
    Issue number2
    Pages (from-to)554 - 566
    Publication statusPublished - 2013


    • Networked visual servo control
    • Real-time image transmission
    • Cloud computing
    • Sending rate scheduling
    • Control with time-varying time delay


    Dive into the research topics of 'Cloud-based Networked Visual Servo Control'. Together they form a unique fingerprint.

    Cite this