Presented by:

Industrial Automation Improvements with Soft Grasp and Hold
January 04, 2018      

Industrial automation is growing quickly in factories, warehouses, and more facilities around the world, thanks to improving capabilities. Just as cheaper sensors have enabled robots to see and move in more dynamic environments, so too has better manipulation sped up the pace and flexibility of robot grasping and piece picking.

Still, improving these capabilities is an ongoing effort. Soft Robotics has focused on doing just that by developing technologies to improve robotic gripping and subsequently the range of applications. We recently spoke with Carl Vause, the CEO of Soft Robotics Inc. in Cambridge, Mass., to understand his product and his outlook on how robots are evolving.

What is unique about your approach to robot grasping?

Robot grasping is based on materials science rather than programming.

Soft Robotics uses materials science for dexterous robot grasping rather than more complicated methods.

Vause: We design and build a new class of soft robotic grippers made entirely of smart elastomeric materials that can manipulate items of varying size, shape, and weight with a single device.

Our fully integrated system can plug and play with existing robots our customers already have in place, which means solutions to difficult automation challenges can be rapidly developed and delivered.

What problems do Soft Robotics’ systems solve, and what was the impetus behind their development?

Unstructured, delicate, and continuously changing tasks used to represent a challenge for automation. Soft Robotics’ technology can manipulate objects that vary in shape, size, and weight, and that are easily damaged.

We saw the most immediate need for this type of technology in food and agriculture, advanced manufacturing, and e-commerce, where automation is needed not only to meet increasing market demand and labor scarcity, but also to manipulate the variable and/or fragile nature of the product.

With Soft Robotics’ actuators, industrial robotics can adaptively handle varied objects with a single device, eliminating the need for tool changes and complex vision requirements. Our technology is now in production, helping our customers to address their automation challenges.

What was the inspiration for your innovations in robot grasping?

Vause: Spun out of the Whitesides Research Group at Harvard University, the design for Soft Robotics technology was inspired by the octopus, a paradigm shift from traditional robotics engineers who are working to address this unmet need with hard linkages, sensors, and servo motors. This inspiration lead to the invention of soft robotic actuators made entirely of polymers that do not require sensors or other electromechanical devices for operation.

A chemist by training, Dr. Whitesides’ invention replaces sensors and processor-intensive numerical computation with material science. By looking to biology for inspiration rather than traditional robotic engineering, we are able to build the fastest and most compliant robotic hands ever produced.

How is the global robotics market maturing? What still needs to happen?

Vause: Today, robotics has made significant headway towards automating repetitive tasks that require little finesse. But tasks that demand manipulation of variable objects and unstructured tasks are still difficult to automate and continue to need human intervention. What makes human beings uniquely suited to solve the challenges and variability presented in a workplace environment? The human brain and the human hand.

With the advent of machine learning and AI, we are training machines how to think like human beings. But replacing the dexterous grasping capabilities of the human hand has always been a challenge. Even if we can train robots to think like humans, with the absence of a human hand, they are unable to perform jobs that require handling of items that vary in size, shape, weight, and fragility.

Even the U.S. government, in its 2016 roadmap for robotics report, labeled “dexterous grasping” as a national imperative to address labor scarcity and the increasing demand for food and fresh produce.

For the robotics market to mature, we need to flip the conversation from robots taking jobs to robots partnering with humans. When we combine our soft robotic gripping systems with machine learning, some amazing things can happen.

Human-supervised automation of tasks like bin picking, sorting, and even harvesting is now a reality. Workers can train robots to pick different kinds of objects so that over time they can move away from high-turnover, dangerous roles and into roles that are safer and more skilled.

Robotics isn’t making human jobs obsolete. It’s making humans more “superhuman” at their jobs. An extension of the brain and the body that makes us faster, more efficient, and finally able to address the increasing demands of the market. The more the industry starts building for these robotic/human-partnered applications, the more I believe we’ll see robotics infiltrating the global workplace.

More on Robot Grasping:

What do you believe makes a particular robotics technology successful? What can make robotics technology unsuccessful?

Vause: Tom Ryden of MassRobotics has a great saying along the lines of, “There is no robotics market; there are problems in markets that can be addressed through robotics.”

I completely agree with this statement. Robots and robotic technologies must be solving an unmet need to be adopted.