For an insect, the ability to sense slight movements can be the difference between life and death. They have been locked for millions of years in an evolutionary arms race with the predators that seek to eat them, and have learned a few tricks along the way.
Take the multi-purpose antennae that allow them to detect and probe nearby objects as an example; the monarch butterfly uses its antennae to orient itself when migrating, and in some species of moth, they double up as flight stabilizers. Other antennae are also sensitive to chemicals: detecting tastes and odors.It’s natural that humans would want to mimic these versatile systems.
Perhaps the best-known example of sensory devices simulating insect features are the biomimetic sensors that draw inspiration from the hairs on cricket legs. These insects can detect predators through minute vibrations and airflows; their filiform, or ‘thread-like’ hairs are called cerci. Cricket cerci have plenty of fascinating features. For a start, there are many of them, and they vary in shape and size, allowing the cricket to distinguish signal from noise in the same way as we might use several sensors to subtract a background pattern of air-flow.
Each of the cerci has an elliptical socket where it attaches to the cricket’s leg, and this means that it prefers to oscillate in a particular direction. For this reason, the cricket’s brain can process not only the speed of the airflow but also which direction the predator is coming from, based on the subset of hairs that are oscillating. There is even some evidence for viscous coupling between the airflow and the cerci themselves.
We can, of course, build sensors that can detect airflow from various directions, but the cerci of crickets are significantly more advanced; they can identify airflow with a sensitivity that’s on the order of 0.03mm/s. This is so sensitive that crickets need a mechanism to respond to airflow that’s due to thermal noise, or Brownian motion of particles; changes in temperature can be picked up as airflow by the crickets. They are functioning on the very edge of the physical ability to pluck a signal from the noise – for them, it can mean the difference between life and death.
A version of this has now been constructed and, what’s more, the scientists noted that they could improve their sensors by more closely mimicking nature regarding the shape of the sensor and its socket. The eventual aim is to create a type of camera that can visualize flows of fluid in real time, with incredibly high sensitivity.
It doesn’t end with crickets. Many insects have fine-tuned means of detecting chemicals. The social system of bumblebees includes components of pheromone detection, and they’re even capable of distinguishing between scents left by bees from their hive or another colony. Bio-inspired chemical sensors use thin chemical films and cantilevers to emulate this effect.
There are several different groups also attempting to develop robot insects – microdrones. Even Wal-Mart has a recent patent for autonomous, flying bees. The military and surveillance applications are obvious, but the flight of insects still holds deep interest for scientists. Many insects can hover, controlling height and flight position with remarkable accuracy. A great deal of insect flight stability and control owes to the unique sensors that insects use.
Trying to create robotics on this scale is immensely challenging: the size and weight of most ordinary components automatically rule them out. The wiring alone can be too heavy for a standard circuit. Teams who work on this, such as the microrobotics team at Harvard University, try to offload as much of the sensory and computational apparatus as possible; but this leaves the drone reliant on excellent communications with a central computer.
Ocelli - Light Sensors
The Harvard team were particularly inspired by the simple light-sensors that insects use – ocelli. The tiny light sensor sits on the head of the robotic insect and allows it to orient itself relative to the horizon. That then allows the robot to estimate its pitch and roll – whether it’s remaining upright and correctly oriented during flight. A feedback loop adjusts the rate at which the robot’s wings flap to enhance stability; this allows the robot to maintain stable flight for far longer than before. Many biologists believe that the ocelli light sensors on insects may serve a similar function.
Of course, the researchers realized that their mimicry doesn’t quite contain all of nature’s secrets; other entomologists note that even when the ocelli are partially obscured, insects can still maintain remarkable flight stability. By constructing bio-inspired sensors, researchers hope to learn more about the natural world as well as how to harness the fruits of evolution.
One of the advantages of looking for biological inspiration is that solutions can be found that run counter to human and engineering intuition. These can include ideas that were, not so long ago, impossible to implement – but due to technological advances and synthesis of new materials, they can be constructed.
One example is the strain sensors insects use. The bodies of some insects are covered in campaniform sensilla, and these sensors have microscopic holes in them. Typically you wouldn’t engineer a component to be riddled with holes, but in the case of a strain sensor, minute changes in strain can be amplified by causing the holes to open or close – which can then be translated into an electrical signal for the insect nerve system. Key to making this design work is the arrangement of hard chitin around the holes, which allows them to bear additional stress without tearing in the course of the insect’s motions.
It is this kind of unique design that the bioengineers seek to mimic. As our ability to micro- and nano-engineer circuitry gets ever more intricate, we will seek to produce machines that merge the best of technology and nature. The solution is to understand the highly efficient systems that many generations of evolution have optimized.