There are some jobs that the human body is not supposed to do. Loading and unloading trucks and shipping containers is a repetitive and arduous task, which is a major reason warehouse injury rates are more than double the national average.
Pickle Robot Company wants its machines to do heavy lifting. The company’s one-armed robot autonomously unloads trailers, lifts boxes weighing up to 50 pounds, and places them on on-board conveyor belts in all types of warehouses.
An homage to the Apple Computer Company, the company’s name hints at the ambitions of its founders AJ Meyer ’09, Ariana Aisenstein ’15, SM ’16, Dan Paluska ’97, SM ’00. The founders want to make the company a technology leader in supply chain automation.
The company’s unloading robots combine generative AI and machine learning algorithms with sensors, cameras, and machine vision software to navigate new environments from day one and improve performance over time. Much of the company’s hardware is sourced from industry partners. For example, you may recognize this arm from a car production line, but you may have never seen one in bright pickle green.
The company is already working with customers such as UPS, Ryobi Tools and Yusen Logistics to help ease the burden on warehouse workers and resolve other supply chain bottlenecks in the process.
“Humans are very good at solving edge cases, robots are not,” Paluska said. “How can robots that specialize in brute-force, repetitive tasks interact with humans to solve more problems? Human bodies and minds are highly malleable, and the way we sense and react to our environments is also highly malleable, and robots aren’t going to replace us anytime soon. But there are a lot of monotonous jobs that we can get rid of.”
Find problems with robots
Meyer and Eisenstein majored in computer science and electrical engineering at MIT, but they didn’t work together until after graduation, when Meyer founded Leaf Labs, a technology consultancy that specializes in building embedded computer systems for robots, cars, satellites, and more.
“My friends from Massachusetts Institute of Technology were running the store,” Mayer recalled, noting that the store is still in operation. “Ari worked there, Dan consulted there, and we worked on some big projects. We were the main software and digital design team for Google’s smartphone, Project Ara, and we worked on a lot of interesting government projects. It was really a lifestyle company for the kids at MIT. But 10 years later, we thought, ‘We didn’t join this company to do consulting. We joined this company to make robots.’
When Meyer graduated in 2009, problems such as robotic dexterity seemed insurmountable. By 2018, the rise of algorithmic approaches such as neural networks had led to major advances in robot manipulation and navigation.
To find out what problems to solve with robots, the founders spoke to people from a variety of industries, including agriculture, food preparation, and hospitality. At some point, they started visiting distribution warehouses with stopwatches to see how long it took employees to complete various tasks.
“In 2018, we went to a UPS warehouse and saw 15 people unloading trucks during the winter night shift,” Meyer recalled. “We talked to everyone, and not a single one worked there more than 90 days. We asked, ‘Why not?'” They laughed at us. They said, ‘Have you tried to do this job before? ”
We found that warehouse turnover is one of the industry’s biggest problems, limiting productivity as managers are constantly hiring, onboarding, and training.
The founders raised seed funding and developed a robot that can sort boxes. It was a simple problem that could leverage technology such as grippers and barcode scanners. The company’s robots eventually worked, but the company wasn’t growing fast enough to make a profit. To make matters worse, the founders were having trouble raising capital.
“We were desperately short on funds,” Meyer recalls. “So we thought, ‘Why would we spend our last dollar on prep work?'”
With funding dwindling, the founders built a proof-of-concept robot that could reliably unload trucks for about 20 seconds at a time and posted a video of it on YouTube. Hundreds of potential customers contacted us. This interest was enough to bring investors back to keep the company afloat.
The company piloted its first unloading system for a year with a customer in the California desert, eliminating the need for human workers to unload shipping containers that can reach temperatures of up to 130 degrees in the summer. Since then, we have expanded our footprint with multiple customers and gained traction among third-party distribution centers across the United States.
The company’s robot arm is manufactured by KUKA, a German industrial robot giant. The robot is mounted on a custom mobile base with an onboard computing system, allowing it to lift and navigate to the dock and autonomously adjust its position within the trailer. Each arm has a suction gripper at the end that attaches to the package and moves it onto a conveyor belt inside the machine.
The company’s robots can pick up boxes ranging in size from 5-inch cubes to 24-by-30-inch boxes. The robot can unload between 400 and 1,500 cases per hour, depending on size and weight. The company fine-tunes pre-trained generative AI models and uses a large number of miniature models to ensure the robots operate smoothly in any environment.
The company is also developing a software platform that can integrate with third-party hardware, from humanoid robots to autonomous forklifts.
“Our immediate product roadmap is loading and unloading,” Meyer says. “But we also want to connect these third-party platforms. Other companies are also looking to connect robots. What does it mean for a robot that unloads a truck to talk to a palletizing robot, or a forklift to talk to an inventory drone? Can they do their jobs faster? I think we’re going to see large-scale networks that will need to coordinate robots and automation across the entire supply chain, from mine to factory to doorstep.”
“Why not us?”
Pickle Robot Company employs approximately 130 people at its offices in Charlestown, Massachusetts. There, standard (albeit green) offices have been replaced by warehouses, where robots can be seen loading boxes onto conveyor belts alongside human workers and production lines.
This summer, Pickle plans to ramp up production of a new version of its system and begin designing a two-armed robot sometime after that.
“My boss at Reef Labs once said to me, ‘No one knows what they’re doing, so why don’t we do it too?'” Eisenstein says. “I carry that with me all the time. I’ve been very fortunate to work with many talented and experienced people over my career, all of whom have unique skill sets and understanding. This is a huge opportunity, and that’s the only way something as difficult as what we’re working on will work.”
The company believes there will be many other robot-related problems with its machines in the future.
“We didn’t start out saying, ‘Let’s load and unload trucks,'” Myers said. “We said, ‘What does it take to have a great robot business?’ Truck unloading is the first chapter. We’ve now built a platform to create the next robots that will start in logistics and eventually support more jobs in manufacturing, retail, and hopefully throughout the supply chain.”
