The WiFi-Bot

Today, WiFi is probably the most important access technology to the internet for end devices. WiFi-based campus networks such as those found at universities, companies and public places are complex, prone to interference and often not optimally configured. The management of such networks is commonly done through central controllers, that receive information from the various access points under their control. It would be better, if these controllers had also access to information from the devices, to see the network from their point of view. Better still would be if those end devices would move around, so that the current network situation can be assed regularly. Of course, doing this with regular means, i.e. personnel carrying around devices, is not a realistic approach as it is too cost intensive. To automate such a task, a robot would probably work well!

The WiFi infrastructure of the University of Applied Sciences Augsburg is also quite complex. There might be misbehaving access points, rouge access points that interfere with our wireless network, and there are places where the WiFi signal strength is weak. Finding these problems would be time consuming and would keep personnel busy for a long time. This project’s goal is to build a robot, that drives through buildings at the university and that scans for these kind of problems autonomously. The potential area of work is quite large. For now, the project will focus on:

  • Reasonably autonomous navigation in buildings
  • Recording of all the important WiFi metrics and central storage of them
  • Displaying the data on the web
  • Simple analysis of the data

There are many bot platforms to choose from. E.g. if climbing stairs is a requirement then the robot probably needs legs, such as Pupper. There are robots that come with a lot of software, that already solves many problems such as simultaneous localization and mapping (SLAM). The TurtleBot3 e.g. is one such platform. At the other end of the specturm, there are the cheap bots, where you need to do everything yourself, such as the Alphabot2. A good compromise between cost, capabilities and features is the iRobot Create 2, a vacuum cleaner bot.

More here, once the project starts. …

Rolf Winter
Rolf Winter
Professor of Computer Networks

My research interests include distributed robotics, mobile computing and programmable matter.