Updated: March 19, 2020 4:37:51 pm
A computer chip, modelled on the human brain, that can smell without a nose? That’s what Intel has achieved with its Loihi, neuromorphic research chip, which will be able to detect the smell of hazardous chemicals. The research was published in Nature Machine Intelligence as part of a joint paper between researchers from Intel Labs and Cornell University.
“We are developing neural algorithms on Loihi that mimic what happens in your brain when you smell something. This work is a prime example of contemporary research at the crossroads of neuroscience and artificial intelligence and demonstrates Loihi’s potential to provide important sensing capabilities that could benefit various industries,” Nabil Imam, senior research scientist in Intel’s Neuromorphic Computing Lab said in a press statement. But what exactly is Loihi and what does this ability to smell out different chemicals mean for artificial intelligence? Here’s a look
What is the breakthrough that Intel is claiming with Loihi?
Intel’s research has shown that Loihi will be able to recognise the smell of 10 hazardous chemicals, including those such as acetone, ammonia, methane. It managed to do this even in the presence of significant noise. The chipset, which is designed to mimic the human brain and its complex neurological systems, achieved this feat of recognising each odour with just a single sample being used to train it. Further, this method did not disrupt the chipset’s previous memories of earlier scents, which were taught to it.
In order to get the chipset to recognise these hazardous smells, the Intel research team relied on a “dataset consisting of the activity of 72 chemical sensors in response to 10 gaseous substances (odours) circulating within a wind tunnel.” The response of these sensors to individual scent was then “transmitted to Loihi where silicon circuits mimicked the circuitry of the brain underlying the sense of smell.”
What are the possible use cases with Loihi’s capabilities?
According to Imam, the idea of an “electronic nose system,” might not be far away. In the future, robots with such neuromorphic chips could be used to monitor and detect hazardous materials or even do quality control in factories. Robots could be used to identify hazardous materials by tracking the smell in different kinds of scenarios.
But there are challenges, given smells can be confusing, and can trip up even the human brain, and these issues will also exist with such computing systems. “These are challenges in olfactory signal recognition that we’re working on and that we hope to solve in the next couple of years before this becomes a product that can solve real-world problems beyond the experimental ones we have demonstrated in the lab,” Imam said.
What is Neuromorphic computing?
Neuromorphic computing is the next level of artificial intelligence, though the idea has been discussed as early as 1940s by computer pioneers such as Alan Turing. It wants to make computers function as closely to the human brain as possible and mimicking that same complex neurological function, but on a chipset.
With neuromorphic computing, the aim is to create a system similar to how neurons fire and interact in the human brain. Just as the human brain can learn on its own, remember a vast amount of information, a neuromorphic chipset attempts to do the same.
The Loihi is Intel’s fifth-generation self-learning neuromorphic research test chip. It was introduced in November 2017. It includes a total of some 130,000 neurons, each of which can communicate with thousands of others, according to the company. The neural network models used on this are much more complex and smarter, than used in other AI-related chipsets, though this technology is still in a research phase.
Intel also announced today that it was able to scale its neuromorphic research system to 100 million neurons capacity with the Pohoiki Springs, which is the company’s most neuromorphic research system. This system basically integrates 768 Loihi neuromorphic research chips inside a chassis the size of five standard servers, according to the company.
📣 The Indian Express is now on Telegram. Click here to join our channel (@indianexpress) and stay updated with the latest headlines
© IE Online Media Services Pvt Ltd