IIT Guwahati unveils brain-inspired AI model for energy-efficient computing

SH²RFSSM combines spiking neural networks with state space modelling to process long data sequences efficiently, with potential applications in edge AI devices

Researchers at the Indian Institute of Technology (IIT) Guwahati have developed a brain-inspired artificial intelligence model that can process long sequences of data with significantly lower estimated energy consumption than many conventional AI approaches, potentially paving the way for more efficient AI deployment on resource-constrained devices.

 

Titled ‘Spiking Heterogeneous Harmonic Resonate-and-Fire State Space Model’ (SH²RFSSM), the AI model combines principles inspired by the event-driven functioning of biological neurons with advanced state space modelling.

 

According to the researchers, the approach is designed to address a key limitation of conventional AI systems, which can become computationally expensive as the length of sequential data grows.

 

 

The model can be used across sectors where continuous data analysis must be performed efficiently. These include wearable health monitoring systems, Internet of Things (IoT) sensors, smart manufacturing, environmental monitoring, autonomous systems and long-term forecasting applications. In these fields, reducing the computational load can extend battery life and enable AI directly on devices without relying heavily on cloud computing.

 

The AI model developed by scientists from IIT Guwahati’s Mehta Family School of Data Science and Artificial Intelligence and SustainAI Lab was widely appreciated at the International Conference on Machine Learning (ICML) 2026 in Seoul, South Korea, one of the leading international conferences in artificial intelligence.

 

Ayon Borthakur, assistant professor at IIT Guwahati, said modern AI systems rely on analysing long streams of sequential data such as health signals from wearable devices, environmental sensor readings, industrial monitoring data and weather or traffic forecasts.

 

"But widely used AI architectures often become computationally expensive as the length of data increases, making them less suitable for battery-powered and resource-constrained devices. The IIT Guwahati team has developed SH²RFSSM, a brain-inspired AI model to address this challenge. The model mimics the event-driven communication of biological neurons," Borthakur said.

 

A distinctive feature of the model is neuronal heterogeneity, where individual artificial neurons are allowed to possess different characteristics rather than behaving identically. This diversity improves the model's ability to capture complex temporal patterns found in real-world sequential data.

 

“Unlike conventional neural networks that continuously process information, spiking neural networks activate only when meaningful events occur, enabling sparse and energy-efficient computation. We combined this principle with advanced state space modelling, allowing the system to learn long-range patterns without the heavy computational cost associated with traditional sequence models," said Kartikay Agrawal, a research scholar.

 

The researchers have evaluated the AI model across 17 benchmark datasets spanning long-range sequence classification, regression, human activity recognition and long-term forecasting. The model delivered performance comparable to state-of-the-art sequence models while demonstrating substantially lower estimated energy consumption, making it promising for edge AI applications.

 

Describing the next steps of the research, Vaishnavi Nagabhushana, another scholar, said the team aims to further explore the model’s potential across real-world applications that involve continuous and long-range data processing.

 

"Our focus will be on improving its efficiency and adaptability so that such AI systems can be deployed more effectively on resource-constrained devices and support practical edge AI applications," she said.