TiHAN at IIT Hyderabad advances autonomous tech for multi-modal transport, tackling safety and regulatory challenges in real-world applications, writes Dr. P Rajalakshmi, a Professor in the Department of Electrical Engineering, CYIENT Chair Professor in Future Communications, Project Director of TiHAN at IIT Hyderabad.
Autonomous technologies are rapidly reshaping the global transportation landscape, revolutionizing public transit, logistics, and urban mobility. Across the world, autonomous systems are moving from experimental stages to practical applications, with countries like the U.S., Europe, and China leading in the adoption of self-driving vehicles, drones, and smart infrastructure. The potential benefits include improved safety, reduced traffic congestion, and increased efficiency. However, significant challenges remain in areas such as infrastructure, regulatory frameworks, and technology readiness.
TiHAN (Technology Innovation Hub on Autonomous Navigation) at the Indian Institute of Technology Hyderabad is playing a crucial role in advancing autonomous technologies. It aims to bridge the gap between research and real-world application by creating innovative solutions for multi-modal transportation, unmanned aerial vehicles (UAVs), and autonomous systems. Funded by the Department of Science and Technology (DST) under the National Mission on Interdisciplinary Cyber-Physical Systems, TiHAN addresses global challenges by developing adaptable and scalable technologies that cater to diverse environments. TiHAN is recognized as a Scientific and Industrial Research Organisation (SIRO) by the Department of Scientific and Industrial Research (DSIR) under the Ministry of Science and Technology, Government of India. A first-of-its-kind state-of-the-art Testbed for Autonomous Navigation (Aerial/Terrestrial) is being developed at TiHAN IIT Hyderabad. The Facilities include Proving Grounds, Test tracks, Mechanical integration facilities like Hangers, Ground control stations, Anti-drone detection systems, State of the art Simulation tools (SIL, MIL, HIL, VIL), Test tracks/circuits, Road Infra –Smart Poles, Intersections, Environment Emulators like Rainfall Simulators, V2X Communications, Drone Runways & Landing area, Control Test centers.
Worldwide Challenges in Advancing Autonomous Navigation
Deploying autonomous technologies on a global scale involves complex challenges, as infrastructure quality, regulatory policies, and public acceptance vary widely across regions. In developed areas with advanced road networks, standardized traffic patterns, and robust digital infrastructure, autonomous vehicles (AVs) benefit from a more predictable and supportive environment, allowing for smoother integration and safer operations. However, in many other regions, road conditions can be poor, with inconsistent signage, irregular lane markings, and varying quality of road surfaces, making it difficult for AV sensors and navigation systems to function reliably. These conditions are further complicated by irregular traffic patterns and unpredictable driver behaviors, which can disrupt the algorithms designed to predict and respond to movement on the road. Additionally, a critical global challenge is the availability of reliable data sources and high-quality mapping, as AVs and unmanned aerial vehicles (UAVs) rely heavily on accurate geographic information and real-time traffic data for effective navigation. In areas with limited data coverage or outdated maps, autonomous systems face difficulties in decision-making and route planning, leading to potential safety risks. The regulatory landscape adds another layer of complexity; while some regions have established clear guidelines for the testing and deployment of autonomous technologies, others lack comprehensive policies, leading to uncertainty and delays in innovation. The absence of standardized global regulations and the differing pace of policy development across countries create barriers to the widespread adoption of AVs, underscoring the need for flexible and adaptable autonomous solutions that can address diverse operational environments.
TiHAN’s Role in Addressing Global Challenges
TiHAN aims to tackle these global challenges by developing technologies that can be adapted to various environments and use cases. It is built inside the campus of Indian Institute of Technology, Hyderabad (IITH). Road infrastructure within the testbed encompasses a diverse range of roadway types, including urban roads, multi-lane highways, collector streets, local streets, and more. Additionally, the testbed offers a wide range of supporting infrastructure for ground vehicle testing, such as signalized intersections, fog/rainfall emulation on roads, and parking zones. The hub’s initiatives span multiple domains, focusing on autonomous ground vehicles, UAVs, and integrated multi-modal transport systems. This facility is essential for validating the safety and performance of autonomous technologies before deployment in more complex environments.
To accelerate the development of autonomous navigation, TiHAN collaborates with industry leaders, government bodies, and international research institutions. These partnerships are crucial in bridging the gap between theoretical research and practical applications. By fostering such collaborations, TiHAN helps bring global advancements in autonomous technologies to India, adapting them to local context. The focus on multi-modal transportation ensures that autonomous solutions can be integrated across various domains.
The regulatory framework developed by TiHAN for Autonomous Vehicles (AVs) emphasizes stringent safety, liability, insurance, privacy, and cybersecurity protocols to support AV deployment. Safety is a core focus, with TiHAN’s testbed at IIT Hyderabad providing a controlled environment where AVs undergo rigorous testing on varied road types and real-world conditions before public road deployment. For liability, TiHAN’s framework advocates for shared accountability among manufacturers, software providers, and users, with black boxes recommended to log incidents and facilitate accident investigations. Insurance provisions are encouraged to cover public and product liability. Additionally, TiHAN’s framework highlights the need for robust cybersecurity measures and data privacy standards, ensuring AVs can securely handle the sensitive data essential for safe operation and public trust.
Sensor fusion Technologies
The TIAND dataset incorporates information gathered from both UAVs and Autonomous Ground Vehicles (AGVs). Sensor data is collected through hardware synchronization from a diverse set of sensors including six Radars (comprising five short-range radars and one long-range radar), one 360° LiDAR, six Cameras offering a 360° Field of View (FOV), and GNSS for precise location data as shown in Figure TiHAN dataset.
TiHAN’s research extends across several cutting-edge fields, aiming to accelerate the adoption of autonomous systems. The hub’s approach includes innovations in AI, machine learning, robotics, and sensor technologies. TiHAN is developing AI algorithms to enhance the perception, decision-making, and planning capabilities of autonomous systems. These algorithms enable vehicles to detect obstacles, identify traffic signals, and navigate complex environments in real-time. A total of 67 test scenarios for Advanced Driver Assistance Systems, covering various real-world conditions for both urban and rural settings has been generated and validated at the TiHAN testbed at IITH.
Further, autonomous systems rely on data from multiple sensors, including cameras, LiDAR, Livox, radar, and GPS. TiHAN’s research in sensor fusion integrates data from these sources to create a comprehensive and accurate representation of the environment.
At the TIHAN testbed, connected vehicle use cases such as pedestrian detection (V2I), pothole detection (V2I), blind spot detection (V2I), weather alert (V2I), emergency brake warning due to stationary/slow-moving vehicles (V2V), and collision warnings with three sub-cases: forward, intersection, and head-on collision warnings (V2V) has been tested and validated. Additionally, lane change warning systems for left/right cut-in scenarios (V2V) using Onboard Units (OBUs), Roadside Units (RSUs), and a connected vehicle stack designed according to IEEE 1609.x standards. This connectivity significantly enhances the situational awareness of autonomous vehicles, enabling safer and more efficient navigation. It supports over-the-air updates, allowing vehicles to receive the latest software upgrades and security patches seamlessly. The deployment of C-V2X technology is supported by the ongoing advancements in 5G networks, which promise to deliver ultra-low latency, high bandwidth, and enhanced reliability.
TiHAN’s multi-modal approach seeks to coordinate various transportation modes through a unified system. For instance, autonomous shuttles can transport passengers within cities while drones deliver goods to more remote locations. These autonomous vehicles utilize deep learning algorithms and neural networks to process the massive amounts of real-time sensor data, enabling them to navigate complex environments while avoiding obstacles and optimizing routes. The e-bikes incorporate electric powertrains, typically powered by lithium-ion battery packs, while featuring smart connectivity modules that enable them to communicate with the broader transportation network.
The shared shuttle services and autonomous cars represent a more complex implementation of self-driving technology, incorporating multiple redundant systems for safety. These vehicles typically employ a combination of GPS, IMU (Inertial Measurement Units), and SLAM (Simultaneous Localization and Mapping) algorithms to maintain precise positioning. They’re equipped with powerful onboard computers running real-time operating systems that can process terabytes of sensor data per hour, making split-second decisions about navigation, obstacle avoidance, and passenger safety.
The UAM (Urban Air Mobility) Based PAV (Personal Air Vehicle) systems represent perhaps the most technologically advanced component, incorporating aerospace-grade navigation systems with redundant flight controls. The autonomous surface and underwater vehicles (ASVs/AUVs) employ specialized sonar systems and underwater communication protocols, using acoustic modems for data transmission and sophisticated pressure sensors for depth management. Real-time data synchronization ensures that all system components maintain updated information about vehicle locations, capacity, and status. The system also incorporates advanced user interfaces, typically through mobile applications, that provide passengers with real-time information and allow them to seamlessly transition between different transportation modes.
Pioneering Edge Computation for Autonomous Systems
TiHAN’s advanced edge computation facility represents a state-of-the-art technological infrastructure that serves as the backbone for autonomous navigation research, powered by a robust combination of Intel Xeon processors and A40 GPU clusters that enable real-time data processing and deep learning model training. This cutting-edge setup facilitates efficient data management through strategic edge computing architecture, where sensor data from various autonomous systems is seamlessly offloaded and processed with minimal latency, enabling fast decision-making capabilities crucial for applications like obstacle detection and path planning. The facility’s computational prowess extends beyond mere processing power, incorporating scalable AI deployment capabilities that support the training and implementation of large-scale artificial intelligence models, which are fundamental to advancing autonomous vehicle technology. Through this integrated approach, the edge computation facility creates a powerful ecosystem where massive datasets can be efficiently processed, complex neural networks can be trained and refined, and real-time autonomous navigation decisions can be executed with unprecedented speed and accuracy.
Summary
TiHAN at IIT Hyderabad is shaping the future of autonomous navigation with comprehensive, end-to-end solutions aimed at global deployment. The state-of-the-art testbed offers real-world testing scenarios, featuring diverse road types, signalized intersections, and weather emulation to validate autonomous ground vehicles, UAVs, and multi-modal transport systems. Addressing global challenges, TiHAN develops adaptable technologies capable of operating in various environments and meeting different regulatory requirements. The TIAND dataset, with synchronized data from radars, LiDAR, cameras, and GNSS, supports the development of robust systems ready for complex navigation tasks. TiHAN’s work is driving smart mobility toward a global destination, ensuring safe, efficient, and adaptable autonomous technology for the future.
Disclaimer: The views expressed by the author are his own and do not necessarily reflect the views of FMM magazine.
Dr. P Rajalakshmi
Professor, Chair Professor & Project Director
TiHAN