Scientists develop miniature anti-collision sensors that can react within two seconds

[ Instrument network instrument development ] A new sensor may help the development of future collision avoidance technologies for drones, robots and autonomous vehicles, inspired by swarms of locusts.

The locusts fly in groups at speeds of 2-3 miles per hour, but they can make evasive movements within a few hundred milliseconds to avoid collisions.

American researchers have developed an electronic version of a special neuron unique to locusts, which allows locusts to respond quickly without consuming too much energy.

The current collision sensors for self-driving cars tend to be large and heavy. The difference is that this insect-inspired detector is very small and can respond within two seconds.

The author of the paper and an engineer at Pennsylvania State University, Saptarshi Das, said: "We are always looking for animals with unusual abilities, animals that do better than humans."

"Insect vision is something that people often use to design automated systems because they can fly without colliding, but then we discovered that locusts are unique."

Unlike other insects, locusts use a single, specialized neuron to avoid collisions-experts call it Lobula Giant Movement Detector, or LGMD for short.

"So we started to study how it works, locusts are incredible. What these creatures can do is very ashamed," Dr. Das added.

The researchers found that LGMD neurons respond to two signals-one from the eye, which is triggered when an approaching locust is seen, and grows as other insects approach-the other It is the relative angular velocity of the locust.

"Because the neuron has two branches, the locust calculates the changes in these two inputs and realizes that something is going to collide, so the avoiding locust changes its direction," said Darsith Jayachandran, an engineer at Penn State University.

Based on the locust's special collision avoidance mechanism, the researchers developed a nano-sized collision sensor using a photodetector made of a single layer of molybdenum sulfide.

They put this device on a tiny programmable memory circuit that uses only a small amount of energy to simulate the neuronal response of a locust in flight.

The team stated that this research and development "is a big step towards smart, low-cost, task-specific, energy-saving and miniaturized collision avoidance systems."

Dr. Das said: "Locusts can only avoid collisions with other locusts, and our equipment can detect potential collisions of various objects at different speeds."

With the completion of the preliminary study, the researchers now need to optimize their equipment to handle situations other than direct collisions-this is exactly what the current setup makes.

"We can't measure every situation. So we developed a numerical model," added Aaryan Oberoi, a paper author and engineer also from the University of Pennsylvania.

"We can also test whether multiple devices work better on the same chip. So far, it seems that a single device is enough. However, a multi-pixel collision detector array can provide collision avoidance in 3D space," he Added.

The full results of this research are published in the journal Nature Electronics.