Driveblocks provides a Physical AI platform for industrial vehicle autonomy, enabling safe and reliable autonomous tasks across agriculture, construction, mining, and defense sectors. The platform integrates perception and sensor-fusion modules designed for harsh, off-road environments and unstructured topologies. It achieves high performance and cost efficiency through fine-tuned AI models derived from a crowdsourced data pool.
Funding
$2.4M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
BKFounders
Product
Problem
Heavy-duty autonomous vehicles often rely on high-definition maps for navigation, which are expensive to create and maintain due to constant environmental changes. This dependence limits the scalability and adaptability of autonomous vehicle deployments in dynamic environments such as mines, farms, and container terminals.
Solution
driveblocks offers a Mapless Autonomy Platform that enables heavy-duty vehicles to perceive their surroundings and navigate autonomously without relying on HD maps. The platform utilizes AI algorithms and sensor fusion to process data from LiDAR, cameras, and other sensors, creating a real-time understanding of the environment. This approach reduces costs associated with mapping and enables vehicles to operate in areas where maps are unavailable or outdated. The modular design allows integration with existing autonomy systems or the development of complete autonomous solutions.
Target Audience
The primary target audience includes OEMs, Tier 1 suppliers, and operators of heavy-duty vehicles in industries such as mining, agriculture, logistics, and container terminals seeking to implement or enhance autonomous capabilities.
Features
- Mapless perception using advanced probabilistic sensor fusion of LiDAR and camera data
- Drivable space detection and corridor identification in real-time
- Object detection and classification using transformer neural networks with attention mechanisms
- All-weather perception for continuous operation in diverse conditions (rain, snow, dust)
- Operation in GPS-denied environments such as tunnels and mines
- Modular software architecture for flexible integration with existing systems
- Perception Development Kit for rapid prototyping and evaluation
- Data-driven classifiers combined with semi-automated training data management and transfer learning