MAN Truck & Bus has launched the “MAN TruckScenes” dataset, the first of its kind from a truck manufacturer, aimed at supporting research and development in autonomous vehicle technology. The dataset, which contains 747 unique driving scenes, is available for free to developers, universities, and research institutions. This move is intended to promote collaboration, accelerate the development of autonomous systems, and contribute to the standardization of data formats in the autonomous driving industry.
The dataset includes data from various sensors, such as four cameras, six lidars, six radars, two inertial measurement units (IMUs), and a high-precision GNSS system. Noteworthy is the inclusion of 4D radar data with 360° coverage, offering a large annotated radar dataset for autonomous truck applications. The focus is on hub-to-hub transportation, with data gathered from German highways, feeder routes, and terminal environments. It also supports machine learning integration, aiding in the development of systems for environment recognition and autonomous driving.
Dr. Frederik Zohm, Member of the Executive Board for Research and Development at MAN Truck & Bus, commented, “Datasets like MAN TruckScenes provide an important resource for data-driven development. With MAN TruckScenes, we are taking the lead to bridge this gap.”
The dataset aims to assist researchers in evaluating and improving autonomous driving systems, particularly for the logistics industry. It covers a wide range of driving scenarios, including varying weather conditions, and provides training, testing, and validation sets to support machine learning model development for autonomous vehicles.
MAN Truck & Bus is involved in several ongoing projects aimed at advancing autonomous truck technology, including the Port of Hamburg Pilot (2018–2020), the ANITA Project (2019–2023), and the ATLAS-L4 Project (since 2022), all of which contribute to the development and integration of autonomous systems in logistics.
