Student Assistant/Project work (m/f/d) - Machine learning based control system for line following cars - in Dresden

The Barkhausen Institute performs research on the latest hardware, software and communication technologies and demonstrates their application in tomorrow's society. Our goal is to build a flexible and trustworthy IoT platform based on reliable electronics and secure communication methods.

To help pursue our research agenda, we are looking for a

Student Assistant/Project work (m/f/d)

• Machine learning based control system for line following cars -

(up to 19 hours/week) 

The project

The cars on our roads have become more and more smart and equipped with sensors. But one key ingredient which is still missing is the communication between the cars. To showcase the importance of this aspect the Barkhausen Institut has developed a demonstrator showing the effects which proper communication between vehicles can have on the traffic on the road.

This demonstrator consists of RC vehicles driving along a predefined track. They keep their trajectory by following a line on the track. Currently this line following is done using a camera mounted on the vehicles and a PID based control loop based on computer vision. Your task is to replace this computer vision application with a machine learning approach. The DonkeyCar[1] Framework is especially developed for such tasks. Your task is to implement a new control system using the DonkeyCar Framework and do a objective comparison of the performance of both approaches, regarding tracking stability, sturdiness against environmental factors, such as different light conditions, and highest possible driving speed.

Your tasks

• Learning about the basics of machine learning in the context of computer vision
• Implementing a custom ML-based control system in the existing demonstrator
• Comparison of the old approach to the new ML approach

We expect the following

• Study of Electrical Engineering or related courses
• Good communication and writing skills in English
• Working with Linux and the command line environment
• Basic understanding of control loops and their performance measures

What we provide

• An open and creative team of enthusiastic scientists
• Close to campus
• Agile working hours and workflows
• Work on future technologies

If you have passion for new technologies and the desire to develop the connected world of tomorrow with us, we are looking for you. 

Life and science are compatible at our institute. We offer a range of recognised family-friendly measures, such as part-time work, mobile working and flexible working hours that take family needs into account.  We also offer support for arriving in Dresden (all formalities, apartment etc.).

Application procedure

Short letter of motivation and curriculum vitae to careers@barkhauseninstitut.org. We only accept complete documents as one single PDF document by e-mail. Please reference “Student Assistant-CoRoLa” in the subject line. Applications that contain other file formats (e.g. Word, Excel) will not be considered in the selection process. Informal inquiries can also be submitted to the same address.

The data submitted by you in the course of the application procedure will be processed by the Barkhausen Institut (BI). BI shall use the data for the purposes of the application procedure. No data shall be passed on to third parties.

[1]https://www.donkeycar.com/