ototo develops a Symbolic AI foundation model designed for safe and fully autonomous driving. This lightweight system explains every driving decision in real-time, offering full transparency for regulatory trust. The architecture enables rapid onboarding to new operational design domains (ODDs) while running efficiently on minimal hardware.
Funding
Funding not disclosed
Founders
Product
Problem
Autonomous vehicles struggle with unpredictable and complex road scenarios, especially those involving interactions between multiple agents like vehicles, pedestrians, and cyclists. Traditional autonomous driving systems often rely on extensive data collection and pre-mapped routes, limiting their ability to handle novel situations and scale to new environments. This can lead to unsafe or inefficient navigation, requiring human intervention.
Solution
Ototo offers a symbolic AI prediction solution designed to enable safe and efficient autonomous driving in complex, real-world scenarios. Their technology utilizes a Symbolic AI-based Foundation Model that mirrors human intuition to predict and plan through dynamic road situations. This approach allows autonomous vehicles to anticipate and react to both visible and hidden road scenarios, even in the presence of multiple interacting agents. Ototo's solution is perception and sensor agnostic, integrating seamlessly with existing hardware and supporting all autonomy levels, from L2+ to L4. The system's explainable decision-making process fosters trust and transparency, eliminating the "black box" effect associated with traditional machine learning models.
Target Audience
Ototo's primary customers are OEMs, Tier 1 and Tier 2 automotive suppliers, and autonomous driving companies seeking to enhance the safety, efficiency, and scalability of their autonomous driving systems.
Features
- Symbolic AI Foundation Model for human-like understanding of complex road situations
- End-to-end solution for prediction and planning in dynamic environments
- Sensor-agnostic technology that integrates with existing hardware setups
- Adaptable algorithm that handles diverse real-world scenarios and supports all autonomy levels (L2+/L3/L4)
- Explainable decision-making process with full transparency and traceability
- Lightweight design that minimizes hardware and computing requirements
- Data privacy focus, operating without storing data from cameras, sensors, or driver behavior