#WeliveAI

Vision becomes reality

Artificial intelligence must also learn

How can an intelligent system be trained to recognize dangerous and unplanned incidents and conditions and act correctly? And how many tests does this system have to go through for people to trust it?

What happens naturally and instinct-driven, must be technologically developed via different artificial intelligence methods and tested over and over again.

Zero error tolerance applies

AI is already in use in many areas nowadays. But human lives are rarely at stake. When it comes to autonomous driving, the safety standards are particularly high and the systems cannot afford to make any mistakes. Since July 2017, this has been ensured by experts from BTC Embedded Systems AG. The subsidiary of BTC AG is a specialist in the field of model-based software development, testing of safety-critical embedded systems and verification strategies for autonomous systems. The experts continuously test the systems according to the highest standards.

Artificial intelligence is the automation intelligent behavior

The requirements for testing autonomous systems are so complex that artificial intelligence is also used here. The part of the development and training of these AI based testing procedures BTC AG takes over. By processing video footage and sensor data with machine learning methods, our Data Scientists answer the question of when an autonomous system has been sufficiently trained to meet current and future safety standards.

Cloud & AI from one source

For the necessary enormous data volumes in the high terabyte range it requires the development and operation of a cloud-based data processing pipeline. Especially for the automotive sector, BTC bundled cloud development & consulting skills. Experienced consultants and developers from the areas of cloud and AI are central contacts for MAN. Another step toward meeting the highest level of security and competence.

„How MAN is on the way to autonomous driving with AI processes.“

Project scope:

Implementation of AI projects in the project context such as development, evaluation and operation of various machine learning models:

  • Among others Convolutional Neural Networks with use of Tensorflow and Keras
  • Architecture consulting
  • Construction and operation of a cloudbased data processing pipeline for data volumes
  • GDPR compliant process for receiving and processing personal video data streams

Benefit:

  • A component on the way to technology leadership
  • Innovation in the field of autonomous driving
  • Central contact persons with cloud and AI experience
Björn Friedrich
Team Data Science und Künstliche Intelligenz