Visit us at the IAA Transportation 2022!

We will launch the first release of our ready-to-drive autonomy platform to enable the software-defined future of the commercial vehicle industry.

driveblocks: Hall 13, Booth C202

The IAA Transportation in Hanover, the meeting spot for the commercial vehicle industry, starts at September, 20th and will bring together vehicle OEMs, fleet operators and startups to discuss the future of trucking and transportation.

driveblocks will present the first version of its ready-to-drive autonomy platform with a focus on hub-to-hub logistics. The algorithms capabilities will be enhanced with each release and extended to further operational design domains, such as container terminals, mines and agriculture applications. The software-based solution enables OEMs to automate their vehicles within weeks instead of years using a pre-configured set of autonomy algorithms. Furthermore, the software modules can be integrated individually with an existing autonomy stack or directly on the sensor hardware to prevent large data transmissions over the network.

There will be several demos available at the booth:

  • Map-free road model in highway situations: The driveblocks road model allows to fuse information from multiple camera images and pointclouds to build a reliable representation of the road ahead. This overcomes the need of precise high-definition maps and allows to scale autonomy applications much faster.

  • Object detection using LIDAR sensors: The driveblocks object detection stack allows to identify the position and size of various traffic participants. In addition, this demo is implemented on top of Apex.OS, a safety-certified fork of ROS 2, developed by our partners Apex.AI. It showcases the parallelization capabilities of the pointcloud preprocessing capabilities and the achieved end-to-end latency improvements.

  • Lane detection using deep-learning enhanced with synthetic datasets: The performance of deep-learning algorithms depends crucially on the quality and diversity of the used dataset and labels for the training procedure. We have teamed up with Applied Intuition, a provider of high-fidelity autonomous vehicle simulation technology, to demonstrate the generation of synthetic training data for lane marking detection and how it enables the development of advanced deep-learning algorithms at driveblocks.

  • Motion planning and prediction in dynamic scenarios: The current release of the driveblocks autonomy platform handles dynamic scenarios in highway situations with ease: A combination of physics- and behavioral prediction for other traffic participants as well as advanced sampling-based motion planning enable semi-hauler trucks to drive in complex traffic situations.