A big leap of mini cheetah

A roping cheetah runs through a rolling field over a sudden gap in rugged terrain. The movement may seem easy, but moving the robot this way is a completely different perspective, MIT News reports.

In recent years, four-legged robots, inspired by the movements of cheetahs and other animals, have made great strides, but lag behind mammalian robots when moving in rapidly changing landscapes.

In these settings, you should use vision to avoid failure. For example, if you can’t see the gap, it’s inevitable to step into it. There are several existing methods for incorporating vision into leg movement, but most of them are really suitable for use in new agile robotic systems, Pullkit Agra, a professor at the Institute for Computer Science and Artificial Intelligence. Gabriel Margolis, a doctoral student in Wal’s lab, says. (CSAIL) At MIT.

Margolis and his collaborators have now developed a system that improves speed and agility as legged robots jump over terrain gaps. The new control system is divided into two parts: it processes real-time input from a video camera mounted on the front of the robot, and it translates that information into instructions on how the robot moves. Researchers tested the system with the MIT Mini Cheetah, a powerful and agile robot built in the lab of Kim Sang-be, a professor of mechanical engineering.

Unlike other methods of controlling a four-legged robot, this two-part system does not require pre-mapping terrain, allowing the robot to move anywhere. In the future, this could allow robots to plunge into the forest on emergency response missions and climb stairs to deliver medicine to the trapped areas of the elderly.

Margolis leads MIT’s Improbable AI Lab and Steven G of the Faculty of Electrical Engineering and Computer Science. And wrote a dissertation with senior author Pulkit Agrawal, an assistant professor of career development at Renee Finn. Professor Kim Sang-bae of the Department of Mechanical Engineering at MIT. I’m Tao Chen and Xiang Fu, MIT graduate students. Other co-authors include Kartik Paigwar, a graduate student at Arizona State University. Kim Dong-hyun, an assistant professor at the University of Massachusetts Amherst. This work will be announced at the Robot Learning Conference next month.

All of it is under control

By using two separate controllers together, this system is especially innovative.

A controller is an algorithm that transforms a robot’s state into a series of actions and follows them. Many blind controllers that do not incorporate vision are robust and effective, but only allow robots to walk on continuous terrain.

Because vision is a very complex sensory input to process, these algorithms cannot process vision efficiently. Systems that incorporate vision usually rely on terrain height maps. Height maps must be pre-built or generated on the fly. This process is usually slow and tends to fail if the height map is incorrect.

To develop the system, researchers took the best elements from these rugged blind controllers and combined them with another module that processes vision in real time.

The robot camera captures a depth image of the next terrain sent to a high-level controller, along with information about the robot’s body condition (joint angle, body orientation, etc.). A high-level controller is a neural network that learns from experience.

The neural network outputs the target trajectory that the second controller uses to calculate the torque for each of the robot’s 12 joints. This low-level controller relies on a series of concise physics equations to describe the movement of the robot, not on neural networks.

This hierarchy, which includes the use of low-level controllers, allows you to limit the robot’s movements and make them work better. This low-level controller uses a well-specified model that can impose constraints. This is usually not possible with learning-based networks, says Margolis.

Teach the network

Researchers have trained high-level controllers using a method of trial and error known as reinforcement learning. They simulated robots running on hundreds of different discontinuous terrains and rewarded successful crossings.

Over time, the algorithm has learned which actions maximize rewards.

We then used a set of wooden boards to build terrain with physical gaps and used a mini cheetah to test the control scheme.

It was definitely fun to work with robots designed in-house by some of MIT’s collaborators. The Mini Cheetah is a great platform as it is modular and mostly made of parts that can be ordered online. If you need a new battery or camera, just order from a regular supplier and you’ll be fine. With a little help from the Sangbaes lab, install it, says Margolis.

Estimating the state of the robot has proved difficult in some cases. Unlike simulations, real-world sensors encounter noise that can accumulate and affect results. Therefore, in some experiments, including precision foot placement, researchers used a motion capture system to measure the true position of the robot.

Their system was superior to other systems that used only one controller, and the mini cheetahs successfully passed 90% of the terrain.

One of the novelties of our system is to coordinate the walking of the robot. If a human is trying to jump over a very wide gap, he may start by running very fast to speed up and then put his feet together to make a very powerful jump over the gap. Similarly, our robots can adjust the timing and duration of foot contact to better traverse terrain, says Margolis.

Jump out of the lab

Researchers have been able to demonstrate that their control schemes work in the lab, but there is still a long way to go before the system can be deployed in the real world, Margolis said. increase.

In the future, we hope to equip robots with more powerful computers so that all calculations can be performed onboard. We also want to improve the robot’s state estimator to eliminate the need for a motion capture system. In addition, we would like to improve the low-level controller to take advantage of the robot’s full range of motion and enhance the high-level controller to function properly in a variety of lighting conditions.

According to Kim, it’s worth noting the flexibility of machine learning methods that can bypass carefully designed intermediate processes (such as state estimation and orbit planning) that model-based methods from centuries ago depended on. Deserves. I am excited about the future of mobile robots with more robust visual processing, specially trained for movement.

This study is partially supported by MIT’s Improbable AI Lab, Biomimetic Robotics Laboratory, NAVER LABS, and DARPA Machine Common Sense Program.

Reissued with permission from MIT News. Please read the original article.

What Are The Main Benefits Of Comparing Car Insurance Quotes Online

LOS ANGELES, CA / ACCESSWIRE / June 24, 2020, / Compare-autoinsurance.Org has launched a new blog post that presents the main benefits of comparing multiple car insurance quotes. For more info and free online quotes, please visit https://compare-autoinsurance.Org/the-advantages-of-comparing-prices-with-car-insurance-quotes-online/ The modern society has numerous technological advantages. One important advantage is the speed at which information is sent and received. With the help of the internet, the shopping habits of many persons have drastically changed. The car insurance industry hasn't remained untouched by these changes. On the internet, drivers can compare insurance prices and find out which sellers have the best offers. View photos The advantages of comparing online car insurance quotes are the following: Online quotes can be obtained from anywhere and at any time. Unlike physical insurance agencies, websites don't have a specific schedule and they are available at any time. Drivers that have busy working schedules, can compare quotes from anywhere and at any time, even at midnight. Multiple choices. Almost all insurance providers, no matter if they are well-known brands or just local insurers, have an online presence. Online quotes will allow policyholders the chance to discover multiple insurance companies and check their prices. Drivers are no longer required to get quotes from just a few known insurance companies. Also, local and regional insurers can provide lower insurance rates for the same services. Accurate insurance estimates. Online quotes can only be accurate if the customers provide accurate and real info about their car models and driving history. Lying about past driving incidents can make the price estimates to be lower, but when dealing with an insurance company lying to them is useless. Usually, insurance companies will do research about a potential customer before granting him coverage. Online quotes can be sorted easily. Although drivers are recommended to not choose a policy just based on its price, drivers can easily sort quotes by insurance price. Using brokerage websites will allow drivers to get quotes from multiple insurers, thus making the comparison faster and easier. For additional info, money-saving tips, and free car insurance quotes, visit https://compare-autoinsurance.Org/ Compare-autoinsurance.Org is an online provider of life, home, health, and auto insurance quotes. This website is unique because it does not simply stick to one kind of insurance provider, but brings the clients the best deals from many different online insurance carriers. In this way, clients have access to offers from multiple carriers all in one place: this website. On this site, customers have access to quotes for insurance plans from various agencies, such as local or nationwide agencies, brand names insurance companies, etc. "Online quotes can easily help drivers obtain better car insurance deals. All they have to do is to complete an online form with accurate and real info, then compare prices", said Russell Rabichev, Marketing Director of Internet Marketing Company. CONTACT: Company Name: Internet Marketing CompanyPerson for contact Name: Gurgu CPhone Number: (818) 359-3898Email: [email protected]: https://compare-autoinsurance.Org/ SOURCE: Compare-autoinsurance.Org View source version on accesswire.Com:https://www.Accesswire.Com/595055/What-Are-The-Main-Benefits-Of-Comparing-Car-Insurance-Quotes-Online View photos