Give robots better movements | MIT News


For most people, the task of identifying an object, picking it up, and placing it elsewhere is trivial. For robots, this requires the latest technologies in artificial intelligence and robotic manipulation.

This is what MIT spinoff company RightHand Robotics has incorporated into its robotic part picking systems, which combine unique gripper designs with artificial intelligence and machine vision to help companies sort products and to place orders.

“If you buy something at the store, you push the cart down the aisle and pick it up yourself. When you order online, there is an equivalent operation in a fulfillment center, ”says Lael Odhner, co-founder of RightHand Robotics ’04, SM ’06, PhD ’09. “The retailer typically has to pick up single items, run them through a scanner, and place them in a sorter or conveyor belt to complete the order. It sounds easy until you imagine tens of thousands of orders per day and over 100,000 unique products stored in a facility the size of 10 or 20 football fields, with the wait for delivery spinning.

RightHand Robotics is helping businesses respond to two megatrends that have transformed retail operations. One is the explosion of e-commerce, which has only accelerated during the Covid-19 pandemic. The other is a shift to just-in-time inventory fulfillment, in which drugstores, grocery stores, and clothing companies restock items based on what was purchased that day or week to improve performance. efficiency.

The robot fleet also collects data that helps RightHand Robotics improve its system over time and allows it to learn new skills, such as smoother or more precise placement. Process and performance data feeds into the company’s fleet management software, which can help customers understand how their inventory moves in the warehouse and identify bottlenecks or quality issues.

“The idea is that instead of looking at just the performance of a single operation, e-commerce companies can change or revise the operational flow throughout the warehouse,” says Odhner. “The goal is to eliminate variability as far upstream as possible, which makes a process simpler and more streamlined.”

Push the limit

Odhner completed his doctorate in the lab of Harry Asada, Ford’s engineering professor at MIT in the Department of Mechanical Engineering, which Odhner says encouraged students to familiarize themselves with robotics research. Colleagues also frequently shared their work at seminars, giving Odhner a comprehensive view of the field.

“Asada is a very well-known robotics researcher, and his early work, and the projects I have worked on with him, is very fundamental to what we do at RightHand Robotics,” said Odhner.

In 2009, Odhner was part of the winning team of the DARPA Autonomous Robotic and Manipulation Challenge. Many competing teams had ties to MIT, and the entire program was ultimately led by former MIT associate professor Gill Pratt. After reaching the semifinals of the MIT 100K competition in 2013 under the name “Manus Robotics”, the team was introduced to Mick Mountz ’87, founder of Kiva Systems (later acquired by Amazon), who encouraged the ‘team to examine applications in the supply chain and logistics.

Today, a significant number of RightHand Robotics employees and leaders come from MIT. MIT researchers also represented many early customers, purchasing components that Odhner’s team had invented during the DARPA program.

“In general, we’ve been so close to MIT that it’s hard to avoid going back,” says Odhner. “It’s kind of a family. You never really leave MIT.

At the heart of RightH and Robotics’ solution is the idea of ​​using machine vision and smart grippers to make part picking robots more adaptable. The combination also limits the amount of training needed to operate the robots, equipping each machine with what the company amounts to hand-eye coordination.

“The technical part of what we do is that we have to look at an unstructured presentation of consumer goods and semantically understand what’s in it,” says Odhner.

RightHand Robotics also uses an end-of-arm tool that combines suction with new under-actuated fingers, which Odhner says gives robots more flexibility than robots that rely solely on suction cups or simple pinch pliers.

“Sometimes it actually helps you have passive degrees of freedom in your hand, passive movements that it can and cannot actively control,” Odhner explains of robots. “Very often these simplify the control task. They take the problems of being heavily over-constrained and make them easy to manage to run a motion planning algorithm.

The data collected by the robots is also used to improve reliability over time and shed light on warehouse operations for customers.

“We can give people information about their inventory, how they store their inventory, how they structure the tasks upstream and downstream of any sample we take,” says Odhner. “We have a really good idea of ​​what may be a source of future problems, and we can let customers know. “

Odhner notes that warehouse management could become a much bigger industry if throughput is improved.

“As consumers increasingly appreciate the ability to shop online, more and more items need to be placed in an increasing number of ‘virtual’ shopping carts. The availability of people near fulfillment centers tends to be a limiting factor for the growth of e-commerce. All of this is really indicative of massive economic inefficiency, and that’s basically what we’re trying to address, ”said Odhner. “We take the less interesting jobs in the warehouse – things like sorter induction, where you just pick up, scan, and put something on a belt all day – and we work to automate those jobs to the point. where you can take your employees and you can direct them to things that will be more directly felt by the customer.

Odhner also claims that more automated distribution centers are offering improved measures to protect worker health and safety, such as ergonomic stations where goods are brought to workers for specialized tasks and increased social distancing. Rather than reducing the number of people employed in a warehouse, he says, “At the end of the day what you want is a system with people working in roles like quality control, supervision. robots. “

Robots made easy

This year, the company introduces the third version of its picking robot, which comes with standardized integration and safety features in an effort to facilitate the deployment of part picking robots for warehouse operators.

“People might not necessarily understand the enormity of our progress in producing this stand-alone system, in terms of ease of integration, configuration, security and reliability, but it’s huge because it means that our robot systems can be shipped almost anywhere in the world. and get up and running with minimal customization, ”says Odhner. “There’s no reason it couldn’t just come in a box or on a pallet and be installed by just anyone. This is our grand vision.

Source link


Comments are closed.