#Germany #Munich #Autonomous Driving #DeepRoute.ai #IAA Mobility

DeepRoute.ai Unveils DeepRoute IO 2.0 at IAA Mobility 2025

In a significant advancement for autonomous driving technology, DeepRoute.ai has introduced its latest platform, DeepRoute IO 2.0, at the IAA Mobility 2025 in Munich. This platform is designed for mass production and features the innovative VLA (Vision-Language-Action) model, which aims to redefine safety and adaptability in autonomous driving across diverse environments.

DeepRoute.ai, recognized as a pioneer in self-driving technology, has already equipped over 100,000 passenger vehicles with its autonomous driving system. The company is also looking to expand its fleet of shared autonomous vehicles (robotaxi) using mass-produced models. This progress highlights not just DeepRoute.ai's production prowess but also its technical maturity in stability, adaptability, and safety standards.

The company's strategic approach is characterized by a two-fold globalization strategy. In Europe, Japan, and South Korea, they leverage their extensive experience in mass production to assist local manufacturers, such as smart, in accelerating their intelligent vehicle offerings. Simultaneously, they provide highly adaptable assisted driving solutions to international automotive manufacturers and joint-ventures, allowing partners to tap into a broader market segment.

Maxwell Zhou, CEO of DeepRoute.ai, articulated, "Our VLA-powered platform combines intelligence and adaptability, making it particularly suited for European road conditions." The DeepRoute IO 2.0 offers exceptional flexibility, supporting both LiDAR-equipped and pure vision configurations, as well as various automotive-grade chips. Its modular architecture enables manufacturers to implement intelligent driving functions across different vehicle models and segments, catering to widespread market demands and facilitating global adoption.

The core of this platform, the VLA model, integrates multimodal vision, language, and actions, enhanced by chain-of-thought reasoning for advanced temporal and causal inference. Compared to traditional end-to-end models, VLA offers superior interpretive capabilities and traceable decision-making processes that foster trust through transparency.

By harnessing the strengths of language models, the VLA incorporates a comprehensive knowledge base and continuous learning ability, guaranteeing rapid adaptation to diverse road systems, traffic signs, and driving cultures. This illustrates its remarkable generalization and resilience, allowing it to function in some of Europe's most challenging driving environments.

Moreover, the VLA model enhances safety with its spatial understanding, logical explanations, OCR (Optical Character Recognition), and voice control features. These capabilities empower the system to proactively respond to high-risk situations such as blind turns, truck blockages, or complex intersections, mimicking human defensive driving maneuvers such as early deceleration and intelligent rerouting.

In merging intelligence, adaptability, and safety, DeepRoute.ai is steering the industry towards a smarter and more reliable future of mobility, shaping a new era of intelligent transport solutions.

About DeepRoute.ai

DeepRoute.ai specializes in artificial intelligence and is dedicated to developing and applying intelligent driving solutions, leading the way in utilizing end-to-end models and VLA architectures in passenger vehicles. Visitors to the IAA Mobility 2025 can explore DeepRoute.ai’s innovations at Hall A2, Booth D25, where a partner will deliver a keynote on September 9 at the Hall B2 Summit Stage. For more information, visit deeproute.ai.