IndraDrive – Autonomous Driver Assistance System designed for Indian highways
Author(s): 1. B. Lakshmipraba, 2.Yash Satish Sail, 3. Shrushti Ramesh Tarvate, 4. Pratiksha Jadhav
Authors Affiliations:
1Professor, Electronics and Telecommunication Engineering, Dr. D Y Patil Institute of Engineering Management and Research, Pune, India
2Student, Electronics and Telecommunication Engineering, Dr. D Y Patil Institute of Engineering Management and Research, Pune, India
3Student, Electronics and Telecommunication Engineering, Dr. D Y Patil Institute of Engineering Management and Research, Pune, India
4Student, Electronics and Telecommunication Engineering, Dr. D Y Patil Institute of Engineering Management and Research, Pune, India
DOIs:10.2015/IJIRMF/202602039 | Paper ID: IJIRMF202602039Indian highways pose uniquely complex challenges for Advanced Driver Assistance Systems (ADAS) due to heterogeneous traffic, weak lane discipline, inconsistent signage, frequent occlusions, and rapid transitions between rural, urban, and intercity environments, making most existing ADAS models—largely trained on foreign datasets such as KITTI, scenes, and BDD100K—unsuitable for Indian conditions, while even recent Indian datasets remain limited in long-tail event coverage and annotation depth. This review systematically examines recent literature on Indian ADAS dataset development, simulation-assisted data generation, supervised learning pipelines, and perception model design, emphasizing the critical need for fully Indian-representative datasets and supervised deep learning architectures trained from scratch using domain-specific data. It highlights the role of simulation platforms like CARLA and SUMO in generating rare and hazardous scenarios, the value of hybrid annotation workflows, and the necessity of evaluation strategies tailored to Indian highways, concluding that reliable ADAS deployment in India can only be achieved through domain-centric dataset engineering and supervised learning pipelines rather than adapting foreign pretrained models, thereby providing a structured reference for future Indian ADAS researchers, dataset creators, and system developers.
B. Lakshmipraba, Yash Satish Sail, Shrushti Ramesh Tarvate, Pratiksha Jadhav (2026); IndraDrive – Autonomous Driver Assistance System designed for Indian highways, International Journal for Innovative Research in Multidisciplinary Field, ISSN(O): 2455-0620, Vol-12, Issue-2, Available on – https://www.ijirmf.com/
- Swathi M, K.V. Suresh, “Automatic Traffic Sign Detection and Recognition: A Review”
- Noor Jannah Zakaria et al., “Lane Detection in Autonomous Vehicles: A Systematic Review”
- Ruohan Li, Yongqi Dong, “Robust Lane Detection Through Self Pre-Training”
- Youcheng Zhang et al., “Deep Learning in Lane Marking Detection: A Survey”
- Hao Ran et al., “FLAMNet: A Flexible Line Anchor Mechanism Network”
- Retaj Yousri et al., “A Deep Learning-Based Benchmarking Framework for Lane Segmentation”
- Nanbin Zhao et al., “A Lane Change Assisting Framework for Large Vehicles”
- Jiping Bi et al., “Lane Detection for Autonomous Driving: Comprehensive Reviews, Current Challenges, and Future Predictions”
- Xinkai Wu et al., “Cooperative Control Research on Emergency Collision Avoidance of Human-Machine Cooperative Driving Vehicles”
- Akash Potdar et al., “PathPilot: OpenCV Based Lane Detection for Safe Navigation”
- Indian ADAS datasets: DriveIndia, IDD, ORDER, iRASTE and open-source dataset/simulation initiatives –
- Rishav Kumar et al., “DriveIndia: An Object Detection Dataset for Diverse Indian Traffic Scenes”
- Adarsh et al., “Advanced Driver Assistance System (ADAS) in India”
- RK Megalingam et al., “Indian traffic sign detection and recognition using deep learning”
- DST, “AI to make roads in India safer to drive” (iRASTE, IDD, ORDER project summaries)