Remote Vehicle Status Monitoring Using Wireless Sensor Networks

Sammanfattning: Status monitoring of vehicles has become more important in recent years, specially for the sake of vehicle testing. Currently real-time vehicle monitoring systems are of more interest to check as many parameters as possible by the means of intelligent and fault auto-detection algorithms and online analysis possibly within the heterogeneous environment.

However previous works have been used wired measuring devices or short range wireless devices like Bluetooth sensors that makes it hard to implement remote monitoring and apply heterogeneity into system. Lack of wireless systems and devices is obvious in this field. Finding an optimized wireless monitoring system is one of the necessities to reduce costs and increase efficiency. therefore we try to investigate possibility of using wireless devices and wireless heterogeneus networks to build up a wireless vehicle monitoring system.

This thesis work tries to show that wireless communications and wireless sensor networks can perform a significant role to improve vehicle status monitoring systems, and also show that integrated data acquired from different types of devices can be used to have better optimization on vehicle monitoring. In the other side integrating WSN with networked embedded system equipped with heterogeneous measurement devices like IMU and CAN-Bus data readers, to acquire raw data from vehicle's outside, inside and engine-side for remote and local visualization enables professionals and analyzers to study vehicle's mechanical and behavioral status remotely.

Project's implementation is based on cross-platform and web based technologies with a local visualizations among a mesh networked local server and remote visualizations through standard communications. In-car WSN and GPS enabled gateways are used to build up a wireless communications between each car and related network.

The results show the absolute possibility of applying wireless systems and heterogeneity into vehicle monitoring as a proof to main question of this thesis work. Final system has been tested with different data acquisition methods on test tracks and test bed at LTU campus. Remote monitoring was basicaly acheived during tests besides data relay to remote server, multirate testabilities and heteregenous communications. Results indicate small improvement requirements on time synchronization between sensor nodes and non-sensor units. Data burst in heterogenous networks are challenging in the sense of packet loss prevention optimizations specially on high rates which was out of the scope of this work. This work had been proved the correctness and possibility of the algorithm and method introduced by Dr. Mikael Nybacka, and became a ground base for a huge amount of ideas for potentialy suitable future works.