Baidu Ventures invests in intelligent traffic solution provider SONDIT

Beijing (Gasgoo)- Chinese intelligent traffic solution provider SONDIT recently announced the successful completion of a multi-million-yuan Series A+ funding round led by Baidu Ventures (BV). 

Photo credit: SONDIT

The substantial investment will be channeled towards pivotal initiatives, including the development of the next-generation high-precision phased-array millimeter-wave traffic surveillance radar, expanding the AI-driven real-time traffic optimization and control system market, enhancing vehicle-road cooperative fusion perception technology, and establishing complementary business structures.

Notably, SONDIT and Baidu have an existing collaborative foundation. SONDIT previously entered into a partnership agreement with Baidu's Intelligent Traffic Business Unit for a smart transportation project. Moreover, Baidu lauded SONDIT as a preferred supplier for Baidu's traffic projects, emphasizing the strategic significance of BV's investment in driving deeper cooperation.

Founded in 2016, SONDIT has been dedicated to addressing urban-level traffic perception and intelligent management challenges. The company boasts comprehensive capabilities in the R&D of phased-array millimeter-wave traffic surveillance radar, encompassing edge computing units, traffic signal control devices, city traffic diagnostic platforms, intelligent highway perception platforms, and real-time intelligent traffic light control systems.

SONDIT has achieved significant breakthroughs, including the autonomous and localized development of traffic surveillance radar technology, overcoming challenges like detection and tracking of dense targets in complex scenarios, noise suppression, road-side and vehicle-road radar compatibility, comprehensive analysis of traffic data, and multi-sensor fusion tracking.

Furthermore, SONDIT's terminal-edge-cloud AI real-time traffic control system, the rToon, bridges the algorithmic and control architecture gap between traffic data and control decisions in urban traffic control. By leveraging deep reinforcement learning and the holographic data acquisition capabilities of radar, coupled with AI algorithms embedded in edge computing units, rToon performs millisecond-level computations and instantaneous signal light timing adjustments based on traffic conditions, achieving localized traffic management optimization and holistic cooperative control.

Up to now, SONDIT’s rToon system has been deployed in various cities, controlling nearly 1,000 intersections in real time.