mit brain.jpg

Singapore researchers are outfitting robots with simulated human brains! With the recent Nobel Prize for the Brain’s Location Code, in 2014, scientists in Singapore are using this discovery to create the next generation of ‘feeling’ robots. These robots are outfitted with simulated human brains that have an innate sense of direction that can make GPS more efficient. What’s next? (image Illustration by John Malta from the MIT Technology Review)

In new and scary science, Singapore researchers are helping robots navigate by giving them a virtual brain that mimics how humans and animals navigate. Often, humans may not memorize roads or streets, but if you’ve been there before, you will probably have some idea of where you are and how to get home. This is because human and animal brains make a mental map of our surroundings on a continual basis – for instance, but using landmarks. We have two types of brain cells that help us navigate called ‘place’ and ‘grid’ cells.


Previously, robots and software would use maps and GPS to navigate, but there are many stories of GPS not always being accurate. In some cases, GPS has tried to navigate folks into lakes and rivers. This new and strange science entirely rethinks computing, in a way. In the research of Haishou Li, at Agency for Science, Technology and Research, they demonstrated how small robots could navigate themselves around an area (without a map) utilizing a digital version of a human brain. The software creates 2D versions of real 3D place and grid cells in order to enable robots to gain a sense of direction and navigation skills that mimic our brain functions.


This was made possible by a discovery of the cells grid and place cells, along with how other cells work together to give animals a sense of location and navigation (sometimes referred to as the location code). Haishou Li and his team of researchers tested their current software and prototype by allowing a robot to roam a 35 square meter office and trying to ensure that the artificial place and grid cells worked the same as real brain cells.


Currently, there is much development that has to happen to ensure this new technology works more efficiently and reliably, however the research is exciting and promising for how computing and robotics may develop in future. This new technology enables more complex systems and networks that current computing can only mimic. A.I. researchers are hoping that this new approach can enable robots to learn and grow – rather than just using code to mimic growth and intelligence.


If Li can refine this new system, navigation systems may become more intuitive and different. But more interesting are the applications for other systems. For instance, what would happen if a search engine was powered by a simulated brain that parses information like humans do – rather than using an algorithm? The caveat is that scientists still have a lot to learn about the brain. The location code is only scratching the surface. It will be interesting to see if, or what scientists will discover about the human brain and how to applying to modern computing in the next decade. Could true Artificial Intelligence happen, or is it just a pipe dream? More interestingly, should it happen at all?

Related Nobel for Brain's Location Code.


See more news at: