Robots have the potential to step into existing workflows and deliver a significant amount of value quickly. But not in all areas of the industry. However, one of the areas where we see multiple robotics companies jump to is handling boxes, specifically in offloading container boxes from the back of a lorry, ideally much quicker than a human in recent times. It is an excellent activity for robots because it showcases its strengths: you may operate in a semi-structured setting, in which speed, energy, and precision are highly prized, and it is not a job for which people are particularly interested or designed to do.
According to Statista: During 2019, around 75,000 logistics robots were sold globally to meet the growing efficiency needs of global trade relations. The value is expected to grow continuously, reaching over 250,000 by 2023.
Dextrous Robotics, a Memphis, Tennessee-based startup headed by Evan Drumwright, is one of the latest approaches to deal with such tasks or jobs. In the beginning, Evan Drumwright was a professor at George Washington University. Then he spent some years at the Toyota Research Institute. In 2019, Drumwright co-founded Dextrous Robotics along with his ex-student, Sam Zapolsky. They have adopted a strategy to do box manipulation without any suction, or indeed without any grippers. Instead, they are utilizing a pair of moving arms that can best be described as chopsticks that take a grip on the boxes.
In an interview with IEEE Spectrum, Evan Drumwright says:
We can pick up basically anything using chopsticks. If you’re good with chopsticks, you can pick up individual grains of rice, and you can pick up things that are relatively large compared to the scale of the chopsticks. Your imagination is about the limit, so wouldn’t it be cool if you had a robot that could manipulate things with chopsticks?
The idea of using moving arms like chopsticks is excellent! But the reason behind using chopsticks is that they can take hold of anything or item, even in huge sizes and quantities, which is highly valuable in constrained spaces where you’ve got significant variations in shapes, sizes, and weights. They are also suitable for manipulation, have the potential to precisely push and place things. Initially, Dextrous was focused on a trailer unloading task; the additional manipulation capacity will make it possible to take more complex manipulation tasks into consideration in the future, like trailer loading tasks that will necessarily take place as often as unloading but are significantly more complicated for automated robots.
Every coin has two sides! The dextrous technique of using chopstick arms has advantages as well as disadvantages. The biggest drawback of the chopstick technique is that using an approach like this is just a lot harder.
“The downside of the chopsticks approach is, as any human will tell you, you need some sophisticated control software to be able to operate. But that’s part of what we bring to the game: not just a clever hardware design, but the software to operate it, too.”
Meanwhile, the usage of suction systems for box handling is very consistent, in what Dextrous has observed so far in this field from other companies. Drumwright explains that the suction system can perform quickly and reliably with limited trials and planning if you have a flat and non-permeable surface (as with most boxes). The suction system, however, has its limitations for manipulation because it is so sticky, implying that it can be difficult and/or time-consuming to do any task with precision. Other problems with suction include the sensitivity to temperature and moisture, its tendency to take all the dirt it possibly can, and the fact that you need to devise the suction arrangement on the basis of the biggest and heaviest items you anticipate to deal with in the future. The last problem in the suction system is that if you want to manipulate smaller objects, you’re left with an enormous array that is much bigger than you would like it to be.
Manipulating small objects in the suction system is a challenge! But it doesn’t mean that the suction system is inferior in any circumstances. Drumwright agrees that a suction system is definitely a feasible option for certain specialized activities. At the same time, chopstick manipulation, if one can make it work, will be more versatile.
Evan Drumwright reveals:
I think there’s a reason that nature has given us hands. Nature knows how to design suction devices—bats have it, octopi have it, frogs have it—and yet we have hands. Why? Hands are a superior instrument. And so, that’s why we’ve gone down this road. I personally believe, based on billions of years of evolution, that there’s a reason that manipulation is superior and that that technology is going to win out.
The secret ingredient to the Dextrous’ Chopstick Robot recipe is its emphasis on stimulation. Dextrous believes that hardware is tough to crack; you want to produce one thing that only works once instead of revising it again and again. It’s definitely impracticable to make it right on the first go, but the closer you can get, the better you can stimulate things in advance.
“What we’ve been able to do is set up our entire planning perception and control system so that it looks exactly like it does when that code runs on the real robot. When we run something on the simulated robot, it agrees with reality about 95 percent of the time, which is frankly unprecedented.”
Dextrous can confidently predict how their system functions even in the most challenging tasks to simulate, like contact and stiction. It is only possible because of very high fidelity hardware modeling, a real-time simulator, and software that can translate straight between sim and real. At the end of the day, Dextrous’ objective is to design a system that can perform more difficult jobs faster than other alternatives.
Wondering why Dextrous uses smooth round chopsticks rather than something with more grip for the manipulation system? Drumwright explains that the benefit of the current design is that it is rotationally symmetric, so you only need five degrees of freedom to control it entirely. It practically means that things can be much more straightforward, checking algorithms become simpler, inverse cinematic algorithms become simpler, and the number of engines in a robot decreases.
Dexterous obtained seed money 18 months ago. Since then, both software and hardware have been used for its box manipulation system and have found time to score an NSF SBIR phase 1 grant. Currently, they are working on chopstick manipulators on two towers that can go parallelly. At the same time, they’re employing the Franka Panda Arms to validate their software. The new equipment will be produced soon, and Dextrous looks forward to delivering paid pilots to real consumers over the coming year.
The global market value of traditional and advanced robotics in logistics is expected to grow between 2018 and 2021. It is forecasted to reach a total value of three billion U.S. dollars in 2021.
The data above suggests that the growth of autonomous mobile robots in the logistics industry will bloom fruitfully to various robot and automation companies, whether it is an established company or a startup.
Dexterous Robotics is a robotics startup founded in 2019, that is working essentially to change how logistics firms operate. It is a team of smart robotics experts trying to transform the logistics industry with robots enabled with artificial intelligence. Their robotic arms with custom AI brains outperform the suction system as chopstick arms don’t tear up heavy or unbalanced boxes, work with speed and precision, operate at all temperatures, and manage porous materials like cloth. Also, it can move a wide variety of items of different shapes and sizes. Their approach could be more versatile than grippers or suction systems.