Enhancing ADAS safety with intelligent, self-calibrating real-time lane detection
drivebuddyAI, a leading provider of AI-powered Advanced Driver Assistance Systems (ADAS) and Driver Monitoring Systems (DMS), has been granted an Indian patent for its invention, “A Real-Time Automotive Assistance System and Method for Lane ROI Detection in Vehicle.”
The patented technology introduces an intelligent, automated framework for lane Region of Interest (ROI) detection that enhances the accuracy and reliability of critical ADAS functions while significantly reducing false alerts across diverse real-world driving environments.
As ADAS adoption accelerates globally, accurately identifying a vehicle’s drivable corridor has become essential for safety features such as Lane Departure Warning (LDW), Forward Collision Warning (FCW), headway monitoring, and driver assistance interventions. Conventional lane detection systems, however, are often affected by variations in camera placement, vehicle architecture, road geometry, environmental conditions, and sensor alignment, resulting in inconsistent performance and inaccurate lane estimation.
drivebuddyAI’s patented solution overcomes these challenges through a self-calibrating framework that automatically optimises lane calculations during new installations, software updates, and system reboots, eliminating the need for manual calibration while ensuring consistent performance.
Key innovations covered under the patent include:
●AI-driven frame validation that selects high-confidence video frames for lane ROI computation.
● Vehicle-specific lane adaptation based on vehicle type and configuration.
● Dual-camera architecture integrating road-facing ADAS with Driver Monitoring Systems (DMS).
● Real-time safety alerts supporting lane departure warning, collision avoidance, and steering assistance.
● Compatibility with Level 2 and Level 3 autonomous driving applications, including Autonomous Emergency Braking (AEB) and advanced path planning.
Unlike conventional approaches that rely on static calibration settings, drivebuddyAI’s technology continuously validates and optimises lane ROI using real-world driving data. This enables reliable performance across inconsistent lane markings, changing road conditions, adverse weather, and diverse vehicle platforms.
The technology has been trained and validated using more than 6.4 billion kilometres of real-world driving data collected across highways, urban roads, rural routes, low-light conditions, adverse weather, and complex mixed-traffic environments. This extensive dataset enables robust lane perception in scenarios where conventional systems often struggle.
Commenting on the patent grant, Nisarg Pandya, Founder & CEO, drivebuddyAI, said:
“The effectiveness of any ADAS system depends on how accurately it understands the drivable space around the vehicle. This patent marks an important step towards self-calibrating ADAS systems capable of delivering more reliable safety alerts with minimal manual intervention.”
The technology is particularly beneficial for commercial fleets and OEMs operating across multiple geographies and vehicle platforms. By automating lane calibration and enabling continuous optimisation through cloud connectivity and over-the-air (OTA) updates, the solution simplifies deployment, reduces maintenance requirements, and helps ensure consistent ADAS performance throughout the vehicle lifecycle.
With this latest grant, drivebuddyAI’s intellectual property portfolio has grown to more than 15 patents, covering innovations across ADAS, DMS, AI perception, driver risk assessment, driver recognition, real-time safety systems, and intelligent mobility technologies. As the automotive industry moves toward connected, software-defined, and increasingly autonomous vehicles, drivebuddyAI continues to advance AI-powered safety technologies designed to make mobility safer, smarter, and more reliable.
