New Perspectives for Innovation

In the last few decades, fast-paced technological innovations have improved our daily lives in ways our grandparents could never have imagined. But to continue to push innovation forward and, equally important, to do so in a sustainable way, we have to think beyond what we know. In other words: we need to look beyond the constraints of our own domain of expertise, combining knowledge from different angles in an interdisciplinary approach.

The progress that is currently being made in the medical field is a case in point. New technologies – wearables, lab-on-chip devices, neuroprobes, and so on – allow us to measure and analyze things that could never be measured before. By combining cutting-edge technology with medical expertise, we gain new insights into how diseases and symptoms develop and how to cure or even prevent them. One recent example to illustrate this kind of interdisciplinary progress is the multi-electrode array (MEA) chip with microfluidic well plates that we have recently developed at imec. Basically, on this tiny chip we can grow several cell structures that accurately mimic a specific human organ, such as heart cells. This way, drugs can be tested faster and more reliably, accelerating their time to market.

But this is not the only area in which we need to look beyond our own perspective. Biology, for instance, can teach us how to advance artificial intelligence and machine learning. Our brain is an unrivalled combination of enormous computing power with low energy consumption, artificial intelligence's ultimate goal. Leveraging our core strength in chip technology, researchers at imec have developed a prototype self-learning chip for neuromorphic computing. In other words, we've developed a chip that learns by imitating the working of our brain. In comparison to existing neuromorphic computing solutions, our chip is more compact and more energy-efficient because it co-optimizes hardware and software.

Our current prototype of the neuromorphic chip is able to compose new music, but in theory the possibilities are endless. Our goal is to further advance both the hardware and software components to achieve very low power, high performance, low cost, and highly miniaturized neuromorphic chips. These could then be applied in a wide range of domains, such as personal health, energy, and traffic management. Integrated into heart-rate sensors, the chip could, for instance, learn to recognize individual electrocardiogram (ECG) patterns and identify significant heart-rate changes that could indicate abnormalities. Another potential application is in prostheses: a chip integrated into a prosthesis could study how an individual patient moves and could then learn to move in harmony with the rest of the body.

These are just two examples to illustrate that technology has not yet reached its limits, but to aspire to what now seems impossible, we need to keep an open mind and collaborate with researchers, innovators, and tech developers from across the globe. Together with other organizations and companies that share this vision on cooperative research – like Hitachi, with whom imec has a long-standing collaboration – we can continue to empower technology as a driving factor for social innovation.