We spoke recently with Carl Vause, the CEO of Soft Robotics, a company in Cambridge, Massachusetts that is making waves with its soft robotic grippers that can manipulate items of varying sizes and weights without the use of sensors or a need for tool changes.
These end-of-arm tooling solutions enable industrial applications that were previously off-limits to automation. They have built a fundamentally new class of robotic grippers that are adaptive, plug and play, repeatable, and reliable.
Q: Can you tell us a little bit about Soft Robotics?
Soft Robotics was born from the research of George Whitesides’ Harvard lab and the discovery of a way to build robotic devices from soft materials such as polymer and rubber. Dr. Whitesides’ revolutionary use of material technology solved several fundamental problems created through more traditional attempts to mimic the human hand in robotics. Traditional robotics had been confined to rigid components such as metals and plastics which needed to be paired with motors and sensors. These systems required a good deal of numerical computation to function and did not have the same dexterity or gentle grasping capabilities.
Robots are units which can accomplish a task repeatedly, without errors. They increase the efficiency of potentially rote tasks such as welding a car, painting objects, and packing items. A major concern in robotics today, however, is the ‘adaptability and agility’ of robots, particularly in advanced manufacturing. Rigid robots, with their heavy programming needs and limited dexterity, are expensive and time consuming.
In contrast, material science has enabled Soft Robotics to perform an array of activities in a highly variable and demanding environment. We can handle meat, processed food, and pastries, pack orders for e-commerce, and pick fruit from a vine, all with reduced programming and no sensors or expensive equipment.
Q: Can you tell us about the applications of your products?
There are three markets that we focus on – food, advanced manufacturing, and logistics.
The food market involves a gamut of products and tasks such as agricultural produce, grocery, bakery automation, handling raw meat, the packaging supply chain, and restaurants. Handling food requires a lot of manual labour that is not readily available. Therefore, food naturally becomes a major market for our gripping solutions.
Our second major market is advanced manufacturing, a ‘high-mix’ environment that requires customisation. Instead of making billions of the same object, advanced manufacturing companies produce smaller quantities of items that are more technically complex and variable. Challenges here include changeover time, skilled labour, and complicated tooling, problems that can be addressed with Soft Robotics.
Our third major specialty is logistics and e-commerce. Online stores are stuffed with a huge variety and quantity of items that need to be picked and packed to address ever-changing customer demands. Agility and ability are important to address the challenges of this environment and Soft Robotics offers both.
Q: Which is more important, your grippers or your software and why?
The answer is both. People tend to gravitate towards our grippers and their special polymer construction. The polymers are non-linear materials that do not behave like traditional robotic devices. It is here that our company’s cutting-edge technology makes the difference.
On the other hand, we’ve built a software stack that enables these systems to operate with one millisecond response time, at 3-4 cycles per second, with precision, repeatedly and reliably, over millions of cycles.
Our software, which enables the smart materials to act like a robotics system, has been a major breakthrough. That’s something that was not done in an academic setting and has been an essential contributor to our success. It’s not just a gripper or the software, however. It is the combination of those two that really makes the thing work.
We can bring in a different process from a different vector, as this system can handle all objects without traditional ‘high information.’ We don’t need to know the weight or hardness of a material. All of that is solved by material science. So, now by bringing in artificial intelligence to enable grafting tasks, the approach becomes entirely different.
Soft Robotics’ grippers are amazing and allow robots to do things that they have never done before. The software then enables speed, reliability, and precision. The AI stack is tailored to solve these problems through the lens of soft materials.
Q: What is unique about Soft Robotics?
Consider ‘dexterous manipulation,’ which has been a fundamental problem in the study of robotic hands with multiple fingers.
Others have been trying to recreate the human hand, an unbelievably complex structure, by using servo motors, sensors, and traditional software stacks. The human hand has developed over millions of years and is backed up by the brain: instead of recreating it, let’s think outside of the box.
Soft Robotics has used material science to solve for the human hand. This technology works, it is real, and it is in production today.
Q: What in your opinion is the future of robotics?
I think human/robotic collaboration is the future. Traditional robotics has been task focused: welding, painting and handling semi-conductor wafers. Humans were separated from robots for security reasons, sometimes by cages.
Today, when the need for automation is so great, we need robots that can work and adapt with humans. When we imagine a human/robotic alliance, we see mobile robots that are agile and adaptable, and robots that work with human to make jobs better. The fun part is that for Soft Robotics, that’s not years in the future, it’s a reality.
Q: What advice can you offer to new start-ups that are going to be entering the robotics sphere?
My number one piece of advice for startups would be to ‘find the problem.’ When we developed Soft Robotics, we had great technology, but we had to find the pain point and then prove that we had the pain reliever. If the pain is acute, people will purchase your product.
If you are going to start a company, it’s the use case which is most vital. We always talk about product markets. I would start with the product market ‘need’ and develop a product around that need.
My second piece of advice would be to spend a lot of time with your market (customers). You’re going to have to test prototypes and get feedback. Companies at all levels get enamored with the technology and develop solutions in their labs without keeping clients in the loop. This is problematic because when the product is taken to market, the customer is unaware of its specifications, uses, and features. Therefore, getting in front of customers, gaining their feedback, and testing in real world applications is critically important.
Q. In conclusion
Large markets such as food and e-commerce were under-represented in automation because robotics didn’t offer the adaptability and agility that they required. Now with Soft Robotics, we are opening doors to bring automation to these very large and meaningful markets around the world.
Robotics News, February 1