Off-road autonomous systems are not as advanced as on-road driver assistance systems, despite high demand in fields such as agriculture, construction, mining, and defence. Factors such as rough terrain, bad weather, thick vegetation, and poor maps make it hard to certify safety and deploy these systems widely.
driveblocks addresses these problems with software that combines camera and LiDAR data to create a real-time 3D view of the environment and guide navigation. This system can operate independently or as a backup to GNSS systems.
The company has just raised €3.5 million in Pre-Series A funding, led by FORWARD.one, with Bayern Kapital, rethink Ventures, and angel investor Joachim Drees also taking part. This brings the total funding to €7.5 million,
The new funding will help driveblocks develop its product, achieve safety certification, and start production agreements with OEMs and system integrators. This supports Europe’s goal of robotics independence and helps address labour shortages.
Building certifiable AI perception for safe, off-road autonomy
Dr Alexander Wischnewski and Dr Stephan Matz founded driveblocks after their success in autonomous racing at the Technical University of Munich, where they won a $1 million Indy Speed Track prize at 280 kph and placed second in a multi-vehicle race in Las Vegas. Their doctoral research highlighted off-road challenges that on-road technologies overlooked.
Wischnewski shares with TFN, “Prior to starting the company, we saw a clear gap for the off-road & commercial vehicle autonomy market, as it has quite different characteristics than the automotive market. After some experimentation with early customers across highway trucking & various off-road verticals, we saw that our technological approach is best positioned to tackle the specific off-road characteristics.”
Its platform combines data from multiple cameras and LiDAR to produce 3D perception using a mix of deep learning and robust algorithms. “Our core focus is camera & LIDAR-based perception, sensor fusion, and local navigation. The platform is purpose-built for off-road applications and therefore excels under challenging conditions,” notes Wischnewski.
What stands out? Robust performance without maps, reliable on-device smarts, and a safety-first approach that scales from pilot to full autonomy. This enables on-device use, quick navigation, and safety certification for any vehicle or application.
Competitors like GUSS and Clearpath Robotics have their niches, but driveblocks is aiming for broader, certifiable impact.
“We have a strong approach to pre-train AI models on a broad database across all verticals and then refine it to a specific vehicle type & application easily. This unlocks several applications in the off-road area and makes it robust & cost-effective at the same time,” explains Wischnewski.
He adds, “Our solution works on production-grade devices – it’s highly optimised for embedded deployment. Its hybrid nature between AI-first and explainable algorithms allows certification in the near term, while enabling us to improve the product over time with more data.”
What about diversity?
On diversity, Wischnewski adds, “We push for a diverse and international team, but we are not tracking statistics as the team still needs to grow to make this meaningful.”
What comes next?
The €3.5M fuels 2026 production pilots with OEMs across ag, construction, and defence, plus new hires in commercial, ops, and R&D right in Munich.
Wischnewski concludes, “The core objective is gradual commercialisation of the technology & AI models. We will focus on use cases with no or very limited safety requirements, prioritising immediate end-user value. Together with our customers, we will bring these into production as soon as possible. “
