Unlike living beings, autonomous vehicles lack ways to filter out useless information, which slows their response time to changes in their environment.
A potential solution? Well, it came to a Purdue research team via none other than those to-go cup lids with the little buttons that are always a joy to push.
“There’s this problem called ‘data drowning.’ Drones cannot use their full flight capability because there is just too much data to process from their sensors, which prevents them from flying safely in certain situations,” said Dr. Andres F. Arrieta, Purdue Associate Professor, School of Mechanical Engineering, with a courtesy appointment in School of Aeronautics and Astronautics Engineering.
Dome-covered surfaces capable of sensing their surroundings would be a step toward enabling a drone’s wings to feel only the most necessary sensory information. Because it only takes minimal force to invert a dome, forces below this threshold would be automatically filtered out.
A specific combination of domes popped up and down at certain parts of the wing, for example, could inform the drone’s control system that the wing is under dangerous pressure. Other potential dome patterns could signify other dangers. Because a dome can adopt only two states — popped up or down — these states can act like zeroes and ones to create patterns for building associative memory, which allows you to construct a much more complete version of the memory based on a partial version.
The team’s work showed that when a certain level of force inverts a dome, embedded sensors surrounding the dome can detect the change. An electrical signal then triggers a memristor to make a record of the force and where it was detected. With each instance of an inverted dome, the metamaterial learns to remember the pattern.
The researchers believe that the metamaterial wouldn’t need to “buffer” to recall information that it stores within itself over time. And since the metamaterial can be manufactured with existing methods, these domes can easily cover a large surface area like a drone’s wing.
Arrieta anticipates that it will be possible to build a drone wing using this material design in the next three to five years.
Here is a Tech Briefs interview, edited for clarity and length, with Arrieta.
Tech Briefs: What were the biggest technical challenges you faced with your work?
Arrieta: The idea is really to try to minimize the number of systems that you want to have and to create materials that can have these relatively complex properties like developing memories and learning. So, the biggest challenge is how can you leverage physical systems and find special material properties that allow you to do that — to try to build memories without conventional hard drives or previously developed types of electronic systems that allow you to do that. And … to embed these properties in the material itself … and also using materials that are flexible, this is another important part.
In our case, we have used mostly problematic materials that are quite flexible. This is by design because we’re interested in creating so-called schemes that could serve as an interface between the physical world and electronic systems such as robots and drones, etc., that interface between the two and are able to help sense, compute, and learn in an efficient way.
Tech Briefs: The team will next test how the material responds to its surroundings based on information it learns from the domes. What kind of testing and research does that entail?
Arrieta: Now we can filter out information that is not important. This is already a relatively big problem in connected systems. We are using our units to find so-called morphological information processing. So we are using the mechanical response to significantly reduce the amount of data that is processed.
Then we process the information and learn it into our memristor. One of the biggest challenges was the production of many memristors, where the memories are actually stored, so that we can couple them to these mechanical units, these domes, and then cover a relatively large area.
At the moment, what we’re trying to do is to be able to have many more of these memristors mounted into our materials. And … we want to be able to learn the pressure distribution over wings of small drones so that we can simplify their control … it could potentially be very advantageous to measure directly how the pressure along the wing is changing. This becomes advantageous when you’re pushing the flight envelope; this is a real problem for military aircraft as they are doing the flight tests and so forth.
One of the main reasons for that is that the control systems, which require signals that are measured and processed to then take positions, are too slow. They introduce a delay, which could be 30 milliseconds, but that is enough to go into a condition that they cannot recover the aircraft from.
Our next step in this technology … we envision that we can significantly reduce these types of delays and also reduce the amount of computational power that is needed to obtain this system state.
Another application we’re looking at is covering robotic arms and legs and hands so that they could learn the environment by touching without having to reprocess all the information or send it to a central computer. If you think about one robot, this is not a big problem; but if you think of millions of robots that are connected — processing all that data can be very complicated. We can already see it now with cloud computing; how the impact of these data centers is incredible in the amount of energy that they consume.
Tech Briefs: Your lab frequently takes inspiration from how spiders and other animals use their anatomical shapes to sense and understand the world around them. What other research has the team conducted?
Arrieta: We have also worked on the morphology of insects’ wings. In particular, we were able to establish the mechanical origins of the stability of the earwig wing … Their wings are one of the wings, if not the wing, with the highest area change from collapse to flying.
The wings are stable; so stable would imply that you don’t need any force to keep them in shape. But people didn’t know why this wing was stable, and this wing has a very interesting origami-like pattern. So we studied it and we were able to determine why. And this happened to be a relatively simple solution from nature: In traditional origami, you do not allow for stretching. If you try to stretch paper, it will tear — and most materials behave like that. Most materials cannot stretch much.
We have also been inspired by the Venus flytrap, a plant that is carnivorous. In particular, we’re focusing on the distribution of properties that the leaf shows to achieve these opening and closing capabilities.
Tech Briefs: Do you have any advice for engineers aiming to bring their ideas to market?
Arrieta: I think that one of the main drivers of our work and the way we think is how can we maintain the complexity of our systems into their design, and how can we create as much multi-functionality from the start? So, many engineering systems is essential.
You start with one system, you optimize it, and you add another system for another property, or maybe two different functions are done by two different systems. And then you have to find a way to make them work together. And, in my view, this leads in many cases to suboptimal solutions — you are always compromising and you're not getting the best of these two systems.
So I would say that for engineers thinking about solutions not to think about solutions that add different systems, but maybe solutions that create systems that can be multifunctional from the start and not be afraid of going for simple solutions.