Dual crickets mounted on, and powered from the iRobot!

Up to now, determining the orientation of the iRobots required the use of the vision based camera system. This was because each iRobot had only one cricked. The solution is simple: mount an additional one!

Additionally to prevent the rise in battery usage, we decided to evaluate the possibility of powering them using the iRobot battery. Using a simple circuit Kees Slinkman designed for us, the crickets now turn on and off with the iRobots and do not require any batteries anymore!

Pictures of the iRobot with two crickets mounted:

Top view of the iRobot

Front/Top view of the iRobot

Waypoint following and avoiding

Movie resulting from half a year's worth of work.

Two robots following waypoints!

The did it! Anton, Graham, Mark and Marko got a first version of the waypoint following algorithm implemented on the setup. Now extending the view of the camera, so the robots don't have to be nudged gently into it's view!

Ready to expand!

This Monday the long awaited USB to Serial adapters arrived. This means that all crickets can finally be connected to the the computer. This should make the localization software more precise!

Now it is just a matter of installing them on the desktop pc and all the net-books. Luckily this is not much work!

Exciting new assignment for the Bachelor research project!

This assignment is available for Bachelor students of Mechanical Engineering at 3mE, TU Delft. Course code: WBTP303, BSc research project (2011-2012 Q1)

In the field of robotics, small relatively inexpensive mobile robots are becoming more and more common. This is a result of the technological advances and development in relatively inexpensive communication, computation and sensing devices. These small robots, can be used in a variety of applications, ranging from search and rescue to surveillance. Although the objective differs from application to application, in general the underlying assumption is that the robots are capable of moving towards points of interest autonomously. Therefore reliable and computationally cheap algorithms have to be implemented and tested to make this assumed capability a reality.

The aim of the project is to develop such a reliable and computationally cheap algorithm that ensures that a mobile robot (or multiple of them) arrives at desired given locations (way-points) with the right orientation. The algorithm(s) have to be implemented and demonstrated on the real test-bed experiment in the Distributed Robotics Lab at DCSC, where six iRobot Create platforms are available.

Once such algorithm is tested and has shown its reliability, free space is given to creativity to design coordinated way-point following methods with all the available mobile robots. So: Let your mind run free and come up with an exciting way to demonstrate your algorithm!

In short, will you join us, and further the research in this exiting area?

There are some demands and requirements for this project. These are:

1. Select and develop a suitable algorithm for way-point following.
2. Design and implement a Matlab interface that uses the information supplied from the available camera system to drive the mobile robot to the desired location.
3. Test and demonstrate the system.

Also, knowledge of Matlab is required. Knowledge of C++ is a plus but not required.

The Group

From the left: Tim, Andrea, Wessel, Jurjen, Tamas, Magnus. Do you want to be part of the group of the Distributed Robotics Lab? mail us: a.simonetto@tudelft.nl

We are on the Slurf

Read our research ideas on the Slurf at this page