A synthetic, insect-inspired “brain” can direct a robotic dog with a fraction of the energy and weight required by traditional methods.
Opteran Technologies, a UK start-up, created a control system that replicates how honeybees and other insects navigate. It is equipped with a computer chip and two cameras that provide a 360-degree view and can be attached to a variety of robots and drones.
The box is only 30 grams in weight and uses less than 3 watts of power. According to Opteran CEO David Rajan, this is a fraction of what typical robot control systems require.
Insects also use visual data when flying, a process known as optic flow. Insects measure both distance and speed at the same time. Researchers are working hard to figure out how insects extract this information just through vision. Nonetheless, algorithms based on this concept demonstrate its viability for future self-driving vehicles.
We will need to improve computing architecture if we are unable to make smaller computer components. The tiny size of insect brains is the inspiration for neuromorphic neural networks. Newer nanoscale technologies imitate brain cell properties to emulate the central nervous system. Newer nanoscale devices simulate the central nervous system by emulating features of brain cells. Insects do not learn how to fill out tax forms or solve hard math problems, but they do efficiently absorb a lot of environmental data.
See Also: Here’s How Robotics Will Transform The Construction Industry
Insects use their sense of smell to detect odorants and other substances. To encode this abstract data, this sensitivity is paired with particular combinations of excitatory/inhibitory brain circuits. Insects compress this information as well so that it may be stored in their brain. This knowledge aids them in classifying different types of scents and responding appropriately. A neuromorphic version of this technique is extremely useful as a classifier that can handle a large amount of data without consuming a lot of resources.
Researchers create sensors and classifiers that incorporate a lot of data without using a lot of energy by stimulating their brains. We’re looking forward to seeing how these technologies evolve over the next few decades.
