Creating an intelligent sensor network for an automotive parking aid system

OData support
Szabó Zoltán
Department of Automation and Applied Informatics

The topic of my dissertation is to design an intelligent sensor network, which installed in vehicles can be used as a parking aid system. I made the project at Robert Bosch ltd. , which possesses great experiences in developing ultrasonic sensors. The primary objective of the project was to develop a service cheaper than the existing one.

The new conception processed in the course of my dissertation transferred the old, centralized system into a new, decentralized network. We managed to reach it by leaving out the central unit ( ECU ) in the old system and the logic was located into the sensors. An 8 bit micro control was built into the sensors. The sensors between them can communicate on a LIN bus , which can provide the necessary noise protection and can be realized easily. The main point of the new conception is that the tasks which were carried out by a 32 bit microcontrol ( in ECU ) want to be achieved by a 8 bit microcontrol working coordinatedly.

In each microcontrol of the intelligent sensor network an embedded operating system was built which provides the right timing of the system and enables to execute parallel processes. The task was to select a small operating system capable of preemptív schedule which supports properly the communication and synchronization between the tasks. Having overviewed the market of embedded operating systems, we chose the one best suits our needs. During the development we paid careful attention to create an architecture which enables to change the operating system simply.

I took part in the project with two other colleagues . My duty was to plan and execute the global system, including determining the critical tasks, dividing the programme into tasks, defining the interfaces which provide communication between the tasks, forming the measuring method, planning the communication between the sensors and the deck computer.


Please sign in to download the files of this thesis.