Robotic Piece-Picking Coming to a Warehouse Near You

An interview with RightHand Robotics’ Co-Founder, Leif Jentoft

This year’s Promat trade show included many interesting warehouse technologies. I found RightHand Robotics (RHR) RightPick solution to be one of the premier exhibits. RHR had its own booth exhibiting its piece picking solution. But perhaps most impressive was the fact that three RHR partners also chose to exhibit RightPick integrated into automated fulfillment processes such as goods-to-picking tending, sorter induction, and automated packaging. While at the RHR exhibit, I had the chance to meet the company’s founders Yaro Tenzer, Leif Jentoft, and Lael Odhner. During our discussion at Promat, I learned that RightHand Robotics’ headquarters is located only 15 miles from my home. I stayed in touch with Yaro and Leif after the Promat show and had the opportunity to visit their corporate office last week and engage in a discussion with Leif about robotics, fulfillment processes, and the future of robotic picking.

Clint: Leif, what was your career path that led you to the founding of RHR?

Leif: My colleagues and I were conducting robotics lab work on how to grab objects in an unstructured environment.  We entered the DARPA Autonomous Robotic Manipulation Challenge. It was taking existing bomb-disposal robots 20 minutes to do a task that people could do in 20 seconds. These robots needed smarter grippers with flexible fingers and sensor feedback.  Our technology met the challenge – won the competition, actually.

We wanted to further develop our technology to take it beyond the lab. After the DARPA challenge, we spoke with hundreds of people in the military, manufacturing, home assistance, prosthetics, and logistics.  We kept hearing the same challenge with picking individual items in a scalable, cost-effective manner to meet the needs of e-commerce fulfillment. Connecting our transformative grasping technology and a huge customer need like piece-picking seemed like an exciting opportunity to start RightHand.

Clint: Earlier this summer I had a discussion with my father -in-law about the new robotics (RightHand) coming to logistics and his response was that robotics have been in industry for years. I am aware of robotic arms in automotive assembly and the use of robotic arms for palletizing in warehouses. How is the problem of piece picking in a warehouse different?

Leif: Robots have indeed been doing great work in factories for decades, where an engineer designs tooling that picks the same item in the same presentation for millions of cycles.  In an unstructured environment like a warehouse, there is incomplete information about the item and a lot of variation – a wider range of items. So the robot needs to be smart enough to pick up different shapes, detect if the grasp was good, know where to place the item, and determine if the item was successfully placed.

Clint: What’s so hard about building robots to handle an unstructured environment like piece picking in a warehouse?

Leif: Flexible grasping has been the holy grail in robotics research for years, but it’s one of the hardest problems because items vary in so many different ways – size, weight, shape, stiffness, etc. Solving it requires deep expertise across machine learning, vision, algorithm optimization, and hardware design. It has been great to learn what the real needs are from the fulfillment professionals on the ground and to bring disruptive technology to bear to solve their problems.

Clint: That true, even a fairly small warehouse has ten thousand different types of items.  How does the RHR solution meet this variability challenge?

Leif: Our solution picks items based on past experience so it can handle the range of inventory.  So when someone in marketing decides to run a promotion on shampoo by increasing the volume per bottle by 30%, our robots can still pick the new version when it arrives unannounced at the warehouse dock. This flexibility also makes a big difference for handling organic items such as produce and deformable items such as apparel.

Clint: That brings up a good point. What types of fulfillment profiles do you think can best benefit from your picking solution, and why? Where is it not as applicable?

Leif: RightPick is designed for piece-picking operations, so it’s typically most useful for items under a few kilograms – grocery, health and beauty, cosmetics, apparel, electronics, bottles, boxes, bags, blister packs, etc.  There are also some items that are challenging for robots – computer cables tend to tangle for example, so they require bagging at receiving first, otherwise they aren’t great to pick.

Clint: In what ways do you plan to improve upon your solution in the future?

Leif: The performance of the RightPick System comes down to what we call the three Rs – the range of items that can be picked, the rate at which they are picked, and the reliability with which the customer gets what they ordered.  The whole engineering team is focused on improving these factors.

RightPick can be deployed across many different workflows – at ProMat this year RightPick was performing a goods-to-picker task at the Intelligrated booth, automated packaging in Accutech’s booth, and sorter induction at Eurosort’s booth, and this will also continue to expand.

Clint: Is your technology currently in use at any customer facilities? Will we be hearing about customer cases in the near future?

Leif: RightPick’s early adopters believe that it’s a competitive advantage. Distribution is shifting towards piece-picking, and they are strategically building automation for a cost-effective scalable solution. These projects are under NDA, but Logistics Viewpoints will be among the first to find out when public use cases become available for general knowledge.

Clint: Thanks Leif. We will keep an eye out for further developments out of RHR.