Skip to main content

Cleaning robot swarm


Visual control of a swarm of cleaning robots


Particularly in the commercial sector (e.g. hotels, offices, retirement homes, hospitals) but also in private use, it is desirable if a floor cleaning by robots in the shortest possible time is completed (noise, accessibility).

Images & Videos


For this purpose, an intelligent, image-based navigation method was developed in which several cleaning robots in the swarm divide the space between each other during the cleaning process. The method uses inexpensive camera sensors with panoramic optics for navigation. It is scalable to any number of robotic robots (even for one robot only), robust against single robot failure and flexible when adding or removing robots. The cleaning is efficient and with low multiple coverage.

The method is based on a number of simple principles, which makes it easy to control the complexity of the control software. Since the swarm robots jointly create a “topological“ map without position information, they do not have to agree on a common coordinate system. The division of the clearance is self- organized by an advanced form of simple “obstacle tracking“ in which each robot drives snugly to real obstacles or to track sections previously driven by any of the robots.  With this ride strategy, the lanes adapt perfectly to the spatial conditions, guaranteeing seamless cleaning with low multiple coverage. The robots communicate only indirectly with each other through the common map construction; a complex planning and coordination with each other is not necessary. A protocol for the exchange of map data via WLAN was developed.


  • Any number of robots
  • Flexible and robust process
  • Controllable software complexity
  • Inexpensive camera sensors


On behalf of the University of Bielefeld, PROvendis is looking for companies that are interested in further development to market readiness. Apart from the cooperation, licenses can be granted for the invention and the intellectual property rights.

PROvendis GmbH

Dipl.-Phys.-Ing. Rolf Klingelberger
+49.208 94105-28
Schloßstr. 11-15
45468 Mülheim an der Ruhr

Development status

Demonstration model

Patent situation

  • DE pending
  • EP pending
  • US pending
  • JP pending


Keywords: Roboter, Schwarmintelligenz, Haushalt, Reinigung, Grünflächen, Robot, collective intelligence, household, cleaning, park area

Offer at Providers website

Contact | Main Office

TransferAllianz e. V.
Christiane Bach-Kaienburg
(Managing Director)

c/o PROvendis GmbH
Schloßstr. 11-15
D-45468 Mülheim an der Ruhr