• autonomous systems
• battery management system
• fail-operational
• fault injection
• FPGA
• lithium batteries
• redundancy

In the last two decades there has been a relevant growth in automotive sector.
Factors such as the care to driver's comfort and safety, together with the interest in reducing pollutant emissions, are increasing the presence of electronic systems inside the vehicle.
This thesis, carried out in the “Autodrive” European project, focuses on a safe handling of Lithium Ion batteries as a main power source for driverless vehicles.
Having to work without the presence of a human supervisor, autonomous vehicles require a high care in fault handling.
On one side is mandatory to have a systematic approach during the design, on the other hand a common verification methodology is required.
The idea followed in this work is to improve reliability of a Battery Management System adopting the Triple Modular Redundancy, extending a commercial BMS.
The extension has been developed onto an FPGA platform.
Compared to a microcontroller approach, an hardware implementation offers several advantages during development and verification, both in terms of flexibility and configurability.
Moreover, the hardware approach has an increased tolerance to single point failures.
Aside hardware, the system was completed with a Graphical User Interface for real time monitoring and its validation.
In fact, both hardware and software are predisposed for fault injection application.
In this way we were able to replicate, in a guided way, the typical battery system faults and confirm the effectiveness of the redundancy approach.