Ripples Of Innovation: Indian Scientist's Team At Oxford Controls Autonomous Robots Using Physical Waves (L: Dr Safeer CK, L:Representative images)

Oxford, UK: In a ground-breaking step towards the future of artificial intelligence, a pioneering team led by an Indian scientist at the University of Oxford’s Department of Physics has demonstrated an innovative method to control autonomous robots using brain-inspired physical waves instead of traditional computer chips.

The study, published in the prestigious journal ‘Nature Communications’, presents a dramatic shift from conventional AI computing. Rather than relying on power-hungry graphics processing units (GPUs) and massive server farms, the researchers turned to the fundamental laws of physics. By harnessing the natural way physical waves spread and interact, the team built a physical AI system capable of processing real-time sensory data and navigating physical space autonomously.

Modern artificial intelligence relies heavily on artificial neural networks running inside vast, power-draining data centres. Training and running these systems requires immense amounts of electricity, while sending data back and forth between a robot and remote servers creates delays – a critical liability for autonomous machines that must make split-second decisions in unpredictable environments.

To overcome this, the Oxford researchers drew inspiration from the human brain, where complex thoughts and actions emerge from populations of interacting neurons. Using a concept known as “reservoir computing,” the team utilised the complex behaviour of waves to process information natively.

“One of the most exciting aspects of this project is that we are bringing together two worlds: fundamental physics and autonomous robotics,” explained Dr Safeer Chenattukuzhiyil, the Keralam-born scientist who led the research team. “Instead of simply studying a physical phenomenon, here we harness the phenomenon itself to process information and make decisions for a robot.”

University of Oxford’s Department of Physics has demonstrated an innovative method to control autonomous robots using brain-inspired physical waves instead of traditional computer chips (Representative image).

Waves are uniquely suited for this task because they naturally travel, collide, and create complex interference patterns across space and time, essentially performing computations through their own natural movement.

To make this invisible computing process visible, the team began with a surprisingly simple setup. They filled a circular metal container with water and dipped electronically controlled actuators into it at precise intervals. These actuators translated incoming environmental data into ripples.

An LED light illuminated the water’s surface, projecting the shifting wave interference patterns onto a screen, captured by a camera. When the system was presented with obstacles, it converted the surrounding environment into wave patterns. The system learned to recognize these distinct ripple patterns with up to 100 per cent accuracy.

When connected to a physical robotic car, the wave patterns directly dictated how the vehicle navigated, allowing it to autonomously steer clear of obstacles. Crucially, the team implemented an “event-driven” approach, meaning the system processed data only when the robot detected a change in its environment. This drastically cut down on energy consumption and significantly improved response times.

While water waves provided a visual proof-of-concept, commercial AI chips demand ultra-fast processing at microscopic scales. The team extended their findings through micromagnetic simulations of “spin waves”—tiny magnetic excitations propagating through magnetic materials. The simulations proved that a spin-wave processor measuring just one micrometre in diameter could carry out the exact same navigation tasks at gigahertz speeds, opening the door for compact, energy-efficient, solid-state AI processors.

Beyond its technological promise, the breakthrough highlights how profound discoveries can arise from creative ingenuity rather than multi-million-pound setups. At a time when frontier research costs are soaring, the team assembled their liquid experimental rig using everyday items: a metal container, water, small motors, an LED, a camera, and standard control electronics.

“For us, this was a reminder that impactful research does not always have to begin with an expensive or complicated experimental system,” said Professor Thorsten Hesjedal, study co-author. “Sometimes, by looking differently at simple physical phenomena and cleverly building experiments using everyday objects around us, we can uncover entirely new possibilities.”

Recognizing the immense market potential, Oxford University Innovation (OUI), the university’s technology transfer partner, has filed a patent application for the underlying computing hardware, paving the way for commercial applications beyond the laboratory.

Dr Safeer Chenattukuzhiyil

The driving force behind this breakthrough is Dr Safeer Chenattukuzhiyil (known professionally as Dr Safeer CK), a Royal Society University Research Fellow who heads the Magnetism for Intelligent Devices (MIND) group based in Oxford’s famous Clarendon Laboratory.

Dr Safeer’s path to the upper echelons of global science is a story of determination and academic excellence. Born in the picturesque village of Mongam in Kerala’s Malappuram district, India, he began his schooling at Mongam Ummul Qura Higher Secondary School before completing his secondary education at Morayur VHM Higher Secondary School.

His passion for science led him to Hansraj College at Delhi University, where he earned his undergraduate degree in Physics. Selected as one of just 12 students across India to receive a prestigious joint scholarship from the Indian Ministry of Science and Technology and the French government, he moved to France to pursue a Master’s degree at Joseph Fourier University.

Dr Safeer went on to complete his PhD in Nanophysics at the SPINTEC laboratory, part of the French Atomic Energy Commission (CEA) in Grenoble. During his doctoral studies, he made pioneering contributions to the concept of “spin-orbit torque memory,” widely recognized as one of the most significant discoveries in spintronics over the last decade. His research excellence earned him a spot as a finalist for the American Physical Society’s Best Ph.D. Presentation Award.

Continuing his research journey across Europe, Dr Safeer secured a prestigious Marie Curie Individual Fellowship at CIC nanoGUNE in San Sebastian, Spain. There, he achieved a milestone demonstration of the spin Hall effect in graphene, establishing key advancements in 2D materials spintronics.

Royal Society is the oldest continuously existing scientist academy, and Britain’s national academy of sciences.

In 2024, Dr Safeer achieved a major career milestone by securing the exalted Royal Society University Research Fellowship—a fellowship awarded by the UK’s national academy of sciences, whose historical ranks include Sir Isaac Newton, Albert Einstein, Srinivasa Ramanujan, and C.V. Raman. Chosen from hundreds of global applicants as one of only five shortlisted candidates at Oxford, he was awarded an individual research grant worth £1.85 million (approx. Rs 19.7 crore).

With this funding, Dr Safeer established the Oxford-MIND group, focusing on brain-inspired physical computing architectures, spintronics, and two-dimensional quantum materials that are only a single atom thick.

To date, Dr Safeer has published 18 research papers in top-tier international journals, including the ‘Nature’ publishing group, delivered over 25 invited lectures across Asia, Europe, and America, and holds multiple international patents for novel magnetic memory schemes and AI hardware. His pioneering work has sparked active collaborations with global technology giants such as Intel and Western Digital.