CDAT™ is Compal’s AI powered perception platform built to deliver consistent detection when visibility becomes unreliable.

By combining infrared sensing with classification, distance estimation, and action prediction, CDAT™ (Classification · Distance · Action · Tracking) helps vehicles recognize and respond to pedestrians and road hazards in the environments where traditional sensors lose confidence. Designed for global platforms and emerging safety requirements, CDAT™  brings dependable perception to real world driving.

CDAT™ classifies a wide range of road users and objects, including:

  • Pedestrians (occluded, static or dynamic)
  • Vehicles (Buses, Trucks, SUVs, Sedans, Semi-Trucks, Light & Heavy Vehicles)
  • 2-wheelers (Motorcycles, Bicycles, e-bikes & Scooters)
  • On road objects
  • Animals

Target Use Cases & Compliance

CDAT™  supports a wide range of ADAS and automated driving applications, from AEB (Automatic Emergency Braking) and FCW (Forward Collision Warning) to more advanced automation levels, including Level 2, 2+, hands-free driving, L3 “eyes-off/hands-off,” and beyond toward L4 autonomy. It is also designed to support regulatory requirements such as FMVSS 127, making it a robust candidate for OEMs targeting safety compliance.

Engineering & Deployment Readiness

CDAT™ is built on a modular, cross-platform perception architecture designed for global vehicle programs. Its models are optimized for efficient deployment across a range of hardware configurations, enabling consistent detection, tracking, behavior prediction, and distance estimation. CDAT™ gives OEMs a ready-to-integrate perception stack that elevates scene understanding and supports proactive safety decisions in real-world conditions.