Harvard has developed the smallest and fastest crawler robot: it can run and jump, only cockroach calves can run and jump, and it can also walk sideways for unprecedented experiments

Produced by Big Data Digest

Source: IEEE

Compilers: Mary, Jie Qiong

What can a robot that is not as tall as cockroaches do?

At ICRA2020, Kaushik Jayaram, Jennifer Shum, Samantha Castellanos, E. Farrell Helbling, and Robert J. Wood from Harvard University and the University of Colorado boulder jointly released "Downscaled Insect-Level Microrobots: HAMR-VI to HAMR-Jr," showing that robots can achieve speeds and small sizes. And it's going to get smaller and smaller!

How fast and small this robot is, please take a look at this coin-sized reptile robot with the digest bacteria!

<h1 class = "pgc-h-arrow-right" > can run and jump, and can also walk sideways</h1>

Developed by the University of Colorado Boulder/Harvard University's Paulson School of Engineering and Applied Sciences (SEAS), HAMR-Jr is coin-sized and one of the smallest and fastest insect-grade robots available today, crawling about 30 centimeters per second.

The last time I visited Harvard Mobile Micro Robot (HAMR) was in 2018, although there was news of "micro robots" at that time, back to now, as far as "insect-level" robots are concerned, the 5 cm long and 3 gram HAMR is still relatively large.

This week at the International Robotic Automation Conference (ICRA), we saw the latest version of HAMR, HAMR-Jr. It is noticeably smaller, weighing only one-tenth the size of a cockroach and as tall as a cockroach's calf.

HamR-Jr is small and capable, and the piezoelectric driver in its body will drive its crawl at a total of 30 centimeters per second, that is, about 200Hz cadence.

Although the driver can work at a faster speed (close to 300Hz), after exceeding 200Hz, the robot will actually slow down, and it has been proven that 200Hz is the best point for resonance, which can make the robot's leg lift and stride length reach a balanced state.

It is worth mentioning that the 200Hz cadence is not slow, and even the insects with the fastest legs cannot reach the 200Hz cadence.

The Australian tiger beetle seems to be the fastest crawling insect in the world on its legs, reaching a speed of 2.5 m/s when chasing prey, but the cadence is only a dozen Hz, which is based on the fastest insect of the same volume size (the fastest speed is measured by body length per second).

If compared like this, it is a tiny mite in California that wins, moving forward at about 200 individuals per second, because the mite is extremely tiny (the size of a sesame grain), which is estimated to be equivalent to 0.25 m/s; but the adult mite can reach a cadence of 135 Hz, which is about the same speed as HAMR-Jr (smaller).

The insect cadence limit may be 100 nhz (microhertz), and when crossing this node, it is necessary to overcome the biological limitations of motor nerve impulses and muscle fiber activity.

The maximum absolute frequency of synchronous contraction of insect muscles is 224hz, such as the tympanic membrane muscles of australian cicadas that make cicada sounds. The term "synchronous" refers to the need for nerve impulses and muscles to work at the same time, but nerve impulses and muscles can also work asynchronously, requiring less active coordination and circumventing some biological control limits.

Some insects use adversarial muscle non-synchronous mechanisms to bend their bodies at extremely high frequencies, often to provide energy for their wings. For example, vampire midges use this mechanism: flapping their wings at a frequency of 1046 Hz, which is similar to the mythical high-frequency speed, which is also severely limited by the wings themselves, and if most of the wing area is removed, the remaining part can also be flapped at a frequency of 2200 Hz.

HAMR-Jr can currently trot, bounce, jump, and walk sideways like a crab. Hamr-Jr still plays steady when its payload reaches its own weight (320 mg), indicating its ability to handle payloads including batteries and sensors. Although it is also desirable to give HAMR-Jr weight loss, the experimental results show that the payload requiring HAMR-Jr is at least 3.5 grams.

In terms of raw speed alone, the biggest performance impact caused by the simultaneous scale-down of the larger HAMR version may be the instability of the HAMR-Jr center of gravity due to the reduction in quality and inertia. One way to solve this problem is to change into a pair of shoes with grip, and you can also adjust the cadence and maximum step size.

< h1 class="pgc-h-arrow-right" > unprecedented experiment</h1>

"We can drive HAMR-Jr leg circulation at more than 200 Hz, a frequency that is almost unprecedented in terrestrial biological systems, and we have no experience with high-cadence motion dynamics experiments." —Kaushik Jayaram, University of Colorado-Boulder

Details of hamr-mini, HAMR-Nano, and HAMR-Atto were obtained through an exchange with Kaushik Jayaram, associate professor of robotics and systems design at the University of Colorado Boulder and first author of the paper, and then presented this evocative research process in a Q&A manner.

Q: When you narrowed down HAMR-VI to HAMR-Jr, what were some of the toughest problems you had?

A: The design and manufacturing process is the easiest part because we don't need to change and debug. So our toughest challenges are mostly from the formation. For example, we are using parts close to the limits of commercial raw materials, and small changes in material (elastic coefficient) or geometric properties (thickness) or glue during assembly can have a significant impact on robot performance. In addition, handling tiny and fragile parts that requires a lot of patience and finger dexterity to combine them under a microscope. Another tricky thing is to ensure athletic performance, mainly because of the 200Hz frequency, which we haven't tried before.

Q: To what extent does the surface that the robot runs affect its performance, and can you elaborate on the potential "surface attachment mechanism" you mentioned in your paper?

A: The ideal running surface of HAMR-Jr should be flat and have good friction, the vertical leg part of HAMR-Jr is moved by about 1 mm, and the surface convexity of more than 1 mm is challenging for it. For larger surface properties, robots will need to "climb" to overcome obstacles. In the past, we have demonstrated negative mechanisms, such as insect-like leg spiny hairs or gecko-like sticky pads, to increase effective friction. We have also demonstrated active attachment mechanisms, such as electrical adhesion, which can be electrically modulated to adhere to metal surfaces, such as the interior of a jet engine, or on a non-conductive substrate, such as under a leaf. We are actively exploring and researching it as a future research direction.

University of Colorado boulder/Paulson School of Engineering and Applied Sciences (SEAS). Researchers say tiny robots like HAMR-Jr help with applications such as search and rescue, industrial inspections, environmental monitoring and drugs.

Q: What aspects of development of a tetherless robot need to be involved?

A: We've shown the power and control of the HAMR-VI (HARM-F) before, which carries all the necessary on-board electronics (about 1 g) and batteries (0.33 g). In fact, HAMR-Jr has a payload capacity of at least 3.5 grams. But we think electronic devices should be able to be made smaller. An article from Robert Wood's lab shows that an electronic device package of about 89 milligrams can drive a miniature executive structure robot like RoboBee, and after trying to reproduce, we got an electronic device package package (<400 mg) close to the weight of HAMR-Jr (about 320 mg) without the need for additional optimizations, which is very exciting!

Q: What are some of the reasons why HAMR-Jr needs to be made smaller? How small can HAMR be?

A: In the face of many real-world applications, we hope that HAMR-Jr will become smaller and more capable. Specifically, the crevices in the rubble of collapsed buildings are usually only a few centimeters long. Similarly, we previously worked with Rolls-Royce: in an engine inspection mission, we found that commercial jet engines had endoscopic ports ranging from 8 to 12 mm in diameter. In robotic surgical applications, the constraints are even smaller, as the largest arteries are only about 8 to 10 millimeters, and making robots of the same scale can also be used as a platform for validating biological hypotheses (especially with regard to insect movement).

I hope to see a more capable HAMR-Jr version that adapts to a cubic centimeter of space in the coming years with full autonomy (power and control).

Q: What areas do you think robots like HAMR-Jr will be able to use?

A: I think the four main scenarios where microbots like HAMR-Jr can have a positive social impact in the coming years are: search and rescue (mechanical flexibility), high-value asset inspection (high autonomy), environmental monitoring (scalability of quantity), and drugs (scalability of size).

The ultra-small robot industry has always been very marketable, especially in some places that humans can't reach, you can use robots to complete difficult work, and I hope that there will be more flexible and capable robots like HAMR-Jr in the future to benefit mankind!