Energy saved, time saved

Earlier this month, the Bar-Ilan Institute of Nanotechnology and Advanced Materials hosted the Israel Society for Plasma Science and Technology Applications (IPSTA) for a conference on innovative studies in the field of material engineering. During a panel on data science concerning new materials, Prof. Hillel Kugler presented the work being conducted in his lab, in the field of biocomputation circuits. “We presented a ‘biological circuit’ unlike any other standard circuit used today, based on electrons and electric currents. A standard electric circuit operates by voltage changes that result from electric currents. To create an electric current, we need to have a large number of electrons move together in a certain direction. We’re replacing the electrons with bioprocessing units (for instance, molecules), each being able to act independently. Each small processing unit can perform a simple, basic function, but since they are independent of one another we can use a large number of processing units simultaneously, and execute complex tasks in a short period of time,” explains master’s student Michelle Aluf-Medina, who is working on her thesis as part of the project. “A circuit based on Biocomputing processing units has two main advantages: one, it saves time—unlike today’s computers, where circuits work in a linear fashion, ours is a parallel circuit that can compute several things simultaneously using different processing units. The second benefit is saving energy, thanks to the computation speed and because the circuit uses biological materials instead of electricity (electrons).”

The study is a collaboration between several European research groups, combining knowledge from various fields, including biology, physics, engineering, and more. Aluf-Medina (31) spearheads the study’s formal verification front. “Formal verification is related not only to biological networks but to many fields in engineering as well. Its goal is to examine design validity before development, to guarantee that logically it should work as intended. We make sure there are no planning errors, in the hopes of avoiding mistakes that cost time, energy, and money,” she specifies. “NBCs themselves are circuits in a maze-like formation, coded to a specific complex problem, and plan them so that the biological materials will be allowed to move in certain routes that together make up and represent the original problem. Then, we introduce into this maze biological material that moves throughout the network in parallel, meaning we don’t have to wait for one to exit in order to let the other entries. We track these materials’ exit routes, and the statistics at which they depart the network provide us with the answer to our question. Of course, like anything biological and physical, we get errors and unwanted behaviors; sometimes the processing units move in unacceptable routes resulting in invalid exits. That is why we also perform statistical analyses and stochastic simulations—so even though there is a certain margin of error in internetwork movement, we can tell if an answer is certainly right or wrong according to a certain number of exiting agents.”

The group’s research currently includes several circuits on which formal validation is performed. This coming May, the validation tool and its simulations will be showcased at the ISCAS conference (under IEEE). At this point, the circuits only provide an answer to a specific question. In the future, discloses Aluf-Medina, the group aspires to create more flexible and modular circuits that can compute different problems with minor modifications. “It’s too soon to talk about applicative uses at this moment, but in the future, our study could interface with medical applications. Generally speaking, interest in this technology is on the rise all around the world. There’s a large research group in Japan, for example, that took these ideas and realized them using lasers and photons; we also have interesting collaborations in progress. The bottom line, we have a biocomputer in our hands. If we can speed up computation time, due to parallelism and energy saving, maybe in the future it could also interface with regular computers.”