Walk to Jail: How NIT Rourkela’s Gait AI Patent is Redefining Security

Author – Ritesh Ranjan: In an era where surveillance systems are expected to be both intelligent and reliable, traditional CCTV-based monitoring continues to show critical limitations. Poor lighting conditions, occlusions, and reliance on manual monitoring often result in missed threats. Addressing these challenges, National Institute of Technology Rourkela (NIT Rourkela) has introduced a breakthrough innovation: a patented AI-powered system that detects unauthorized individuals using thermal imaging and gait recognition.
With Patent No. 580748, this system represents a major leap in automated security. Designed to function seamlessly in low-light and high-risk environments, it combines thermal sensing with advanced machine learning models to identify intruders based on how they walk.

The Problem with Traditional Surveillance
Conventional CCTV systems have long been the backbone of security infrastructure. However, they come with inherent flaws:
- Heavy dependence on lighting conditions
- Reduced accuracy in crowded or obstructed environments
- Manual monitoring inefficiencies
- Vulnerability to disguises and facial concealment
Recent studies highlight that a significant percentage of threats go undetected due to these limitations. This gap has created an urgent need for more robust, intelligent, and autonomous systems.
The Innovation: Thermal Imaging Meets Gait Recognition
The patented system developed at NIT Rourkela integrates two powerful technologies:
1. Thermal Imaging
Thermal cameras detect infrared radiation emitted by human bodies, enabling identification even in complete darkness, fog, or smoke. Unlike visible-light cameras, they are not affected by environmental lighting conditions.

2. Gait Recognition
Gait recognition analyzes the unique walking pattern of individuals. Every person has a distinct stride influenced by body structure, posture, and movement dynamics. This system evaluates nearly 20 different gait parameters using machine learning models such as LSTM (Long Short-Term Memory networks).
The combination ensures high accuracy in identifying individuals without relying on facial data, making it both effective and privacy-conscious.
How the System Works
The system operates through a multi-stage process:

Detection
Thermal cameras continuously scan the environment to identify human presence based on heat signatures. This reduces background noise and false positives.
Tracking
Multiple cameras coordinate through a centralized server to track the movement of individuals across different zones.
Identification
The system compares captured gait patterns with a pre-existing database. If no match is found, the individual is flagged as unauthorized.
Alert and Logging
Real-time alerts are generated for security personnel. Temporary data is stored for analysis and forensic use, while ensuring efficient data management.

This entire process is automated, minimizing human intervention and improving response time.
Development and Research Team
The innovation is led by Professor Samit Ari from the Electronics and Communication Engineering department at NIT Rourkela. The project also includes contributions from researcher Mohammad Iman Junaid and MTech alumni Narayan Prasad Sharma and Irshad Ali.
The research received funding support from the Science and Engineering Research Board (SERB), highlighting its national significance and potential for large-scale deployment.
Key Advantages of the Technology
Works in Low-Light Conditions
Thermal imaging ensures uninterrupted surveillance regardless of lighting conditions.
Non-Intrusive Identification
Unlike facial recognition systems, gait analysis does not require capturing sensitive biometric data, making it more privacy-friendly.
High Accuracy in Crowded Environments
Gait-based identification remains effective even when faces are obscured or individuals are partially hidden.
Cost-Effective Deployment
The prototype costs approximately ₹1.9 lakh, with potential reduction to ₹50,000 per unit at scale, making it viable for widespread adoption.
Scalability
The system can be expanded from small setups to large industrial or urban infrastructures.
Real-World Applications
Educational Campuses
Pilot testing at NIT Rourkela has reportedly reduced unauthorized access significantly. Similar institutions are exploring adoption for campus security.
Industrial Facilities
Factories and research labs, especially in high-value sectors, can benefit from continuous monitoring and automated threat detection.
Defense and Border Security
The system’s ability to function in low visibility conditions makes it ideal for border surveillance and sensitive military installations.
Smart Cities and Transportation
Airports, railway stations, and urban surveillance systems can integrate this technology for enhanced safety, particularly in challenging weather conditions.
Market Potential and Industry Impact
India’s security and surveillance market is expanding rapidly, with increasing demand for AI-driven solutions. The introduction of an affordable and efficient system like this could transform how institutions approach security.
The system also aligns with global trends where countries are exploring non-invasive biometric technologies. By focusing on gait rather than facial data, it addresses both operational efficiency and privacy concerns.
Additionally, the cost advantage positions it as a competitive solution not only in India but also in emerging markets worldwide.
NIT Rourkela: Driving Innovation
NIT Rourkela continues to establish itself as a hub for cutting-edge research and innovation. With a strong focus on artificial intelligence, computer vision, and applied engineering, the institute has consistently contributed to technological advancements with real-world applications.
Its growing portfolio of patents and industry collaborations reflects a commitment to solving critical challenges through research-driven solutions.
The Road Ahead
The next phase for this technology involves industry partnerships and commercialization. Potential developments include:
- Integration with drone surveillance systems
- Deployment in smart city infrastructure
- Collaboration with defense organizations
- Development of portable and app-based solutions
As adoption increases, the system could become a standard component of modern security frameworks.
Conclusion
The patented gait recognition and thermal imaging system from NIT Rourkela represents a significant breakthrough in surveillance technology. By overcoming the limitations of traditional CCTV systems, it offers a smarter, more reliable, and privacy-conscious approach to security.
As institutions and industries look for advanced solutions to safeguard assets and people, innovations like this are set to play a crucial role in shaping the future of security in India and beyond.
FAQs
1. What is gait recognition technology?
Gait recognition is a biometric method that identifies individuals based on their walking patterns, which are unique to each person.
2. How does thermal imaging improve security systems?
Thermal imaging detects heat signatures, allowing surveillance systems to function effectively in darkness, fog, or smoke.
3. Is this system better than traditional CCTV?
Yes, it addresses key limitations of CCTV, such as poor visibility and reliance on manual monitoring, offering automated and accurate detection.
4. Where can this technology be used?
It can be deployed in campuses, factories, defense installations, airports, and smart city infrastructure.
5. Is gait recognition safe for privacy?
Compared to facial recognition, gait analysis is less intrusive as it does not rely on capturing identifiable facial features.





