State-of-the-art chip designs developed in the lab of Shreyas Sen, an Associate Professor at Purdue University, could offer groundbreaking solutions to the energy constraints plaguing today’s wearable devices. These energy limitations have long restricted the potential of wearable gadgets to handle complex AI tasks, thereby affecting their utility in critical areas such as healthcare, where continuous, real-time monitoring is vital. Imagine a wearable electrocardiogram that could detect heart irregularities without relying on constant internet connectivity. Such a device could save lives, even in areas with poor signal reception.
Sen, who holds the Elmore Associate Professorship in Electrical and Computer Engineering at Purdue, has drawn inspiration from the human body’s own nervous system to create chips for wearable devices that could function offline and require significantly less frequent charging. This fascinating approach aims to overcome one of AI’s biggest hurdles: its high energy consumption. Currently, AI algorithms operate within massive data centers, which consume tremendous amounts of power to process information. Wearable devices must continually connect to these data centers via the internet, limiting their functionality and battery life.
The tech industry has been pushing toward edge technology – chips capable of running AI algorithms directly on devices such as smartphones and smartwatches. Sen’s lab is at the forefront of this research, creating custom chips designed to handle complex AI tasks with minimal energy expenditure, effectively mimicking the cooperative effort between the human brain and peripheral nervous system.
Chips designed by Sen’s team have shown promising results. For instance, one of their integrated circuits, inspired by the way the human body transmits information, utilizes electro-quasistatic signals to transfer data up to ten times faster than conventional Bluetooth, while consuming a hundred times less energy per bit. This breakthrough, presented at the 2022 IEEE International Solid-State Circuits Conference (ISSCC), potentially revolutionizes data transfer between wearable devices, paving the way for more efficient and powerful AI applications.
Sen’s innovation doesn’t stop at wearable devices. His research ventures into brain implants, demonstrating the first-ever use of electro-quasistatic signals for communication within the brain. These preliminary results signify a monumental leap towards smarter, more efficient health monitoring systems and other AI-driven applications.
In practical terms, this means a future where wearable devices like smartwatches or ECG patches won’t need constant internet access to function effectively. For example, a patch with an in-sensor analytics chip on your chest could alert you to irregular heart activity without requiring a stable internet connection. Not only would this ensure constant monitoring, but it would also alleviate the need for frequent recharging.
The culmination of these advancements is embodied in Ixana, a startup co-founded by Sen and his team, dedicated to commercializing these sophisticated designs. Their flagship product, Wi-R, demonstrates the extraordinary capability of these new chips. Developed to capitalize on the human body’s own information transmission methods, Wi-R showcases how these chips could transform wearable tech. Debuted at the CES tech show in Las Vegas, Wi-R has already garnered significant attention, winning two CES innovation awards and making it to the Electronic Engineering Times’ list of 100 startups to watch in the silicon industry for 2024.
With these efforts, Sen and his collaborators at Purdue are not just improving wearable devices; they are redefining the way AI can be integrated into everyday objects, drastically reducing energy consumption while expanding functionality. This approach could potentially unlock new, unimaginable applications for wearable technology, further bridging the gap between sophisticated AI and everyday consumer needs.
Purdue University continues to be a hub of cutting-edge research and innovation, supporting ventures like Sen’s with the aim of pushing the boundaries of what’s possible. This relentless pursuit of advancing AI technology in a sustainable, energy-efficient manner exemplifies Purdue’s commitment to pioneering research and its practical applications in improving human lives.