From lights-out manufacturing to lifesaving treatments

In the medical industry, automation and robotics applications range from attention-grabbing, with surgeons leveraging robots to perform difficult, risky procedures, to more low profile, such as automated programs that streamline administrative tasks. Then there’s the automation taking place outside of hospitals and clinics, in themanufacturing facilities fabricating the surgical tools, implants, and numerous other pieces crucial to patient care. Whether highly visible at the point of care or more under the radar in factories, robotics and automation have become critical to medtech.

Robots in the clinic and on the factory floor

Robot-assisted surgery is one of the most well-known applications of automation in hospitals, enabling less invasive and risky procedures, faster recovery times, and greater precision and range of motion for the surgeon. Robots also make up much of the fabric of a hospital’s day-to-day operations, however – tracking and maintaining inventory, cleaning and sanitizing, and providing sterile transportation of vital supplies throughout the facility.

“When companies adopt automation, they’re often adopting it because they can’t find people to do the dull, dirty, dangerous jobs that people really don’t want to do or shouldn’t be doing,” says Jeff Burnstein, president of the Association for Advancing Automation (A3). A3 is the largest automation trade association in North America, with the goal of promoting and educating about opportunities for robotic implementation in a variety of industries.

Burnstein emphasizes people shouldn’t be concerned about robots taking their jobs – on the contrary, the machines offer a boost for employees looking to advance.

“The technologies are there to augment people,” he continues. “So people, instead of doing jobs like lifting heavy things all day, can do jobs where they oversee the technology, oversee the robots, so they have better jobs that add more value to the company, and oftentimes those are higher paying and certainly safer jobs.”

That’s certainly the case on the manufacturing side of medtech. Many factory floors are running more smoothly and quickly thanks to autonomous mobile robots (AMRs) – the hottest trend in manufacturing right now, according to Corey Ryan, director of medical robotics at KUKA Robotics Corp. These robots can take on both burdensome physical tasks – moving carts, pallets, boxes, etc. – and tricky mental ones.

“If you have 10 AMRs, you can optimize them to be constantly choosing the closest unit, constantly looking for availability of any special features and any special components that are required, and you can prioritize high-demand parts or urgent deliveries over anything else,” Ryan says. “And it all happens automatically, versus having somebody who’s paying attention to the delivery schedule and timing of the manufacturing.”

Small batch manufacturing is one area in which automation has potential to advance quickly, and it’s the current focus of many cobot integrators. Cobots can be easily deployed and can reuse code in different applications – and when they’re combined with AMR technology, they can move from task to task without human intervention, opening up new possibilities for lights-out manufacturing.

Where artificial and human intelligence converge

Automation is an umbrella term for a vast ecosystem of robotic technologies – on one end are the robots completing physical tasks, and on the other end are the machine vision and artificial intelligence (AI) applications. In clinical settings, this may include checking pharmaceutical vial labels for accuracy or analyzing patient scans for more precise diagnoses and treatment courses. In manufacturing facilities, 3D bin picking is what Ryan calls “the Holy Grail of manufacturing.” This involves training AI to select the right part from a bin full of varying parts, which is currently still a challenge for machine vision. It’s a challenge for human vision as well, as even the sharpest eyes can get crossed when scanning a vast array of similar parts.

Bringing automation into a field with the traceability needs and regulatory demands of medical manufacturing is tricky. AI can be an asset in tracking the design and production of medical devices, eliminating the potential for risky human error, but it still lacks the capability to reason or adapt to unexpected variables.

“We talk about AI taking over some of the decision making, but there’s a real problem with AI in that a lot of it’s black box,” Ryan points out. “There are a lot of models that’ll tell you what the best answer is, but can’t tell you why that’s the best answer. That’s a problem, because the reason for the solution is sometimes as important as the solution itself.”

Thus, humans are still a vital element; the effective combination of a human’s critical thinking and reasoning skills and a robot’s precision and repeatability are why cobots have become so common on the factory floor. Ryan echoes Burnstein’s words about the “dull, dirty, dangerous” jobs robots can take on but stresses the limitations these machines have without proper historical data provided by humans. Many manufacturers, while they have plenty of data about the amount of time it takes to produce a part, for example, don’t have the process data a robot needs to perform that production: what motions did humans use to make the part, and how can those motions be optimized?

“I think as we start to break manufacturing processes down into smaller steps that AI can handle, we’ll get interesting results,” Ryan says. “We just have to really be careful that we validate the reasoning behind them, and validate that they’re actually optimized in a way that works best for the whole.”

This is one reason it’s so urgent to educate the current and future workforce on how to design, program, and work with robots, and why many experts say expanding automation will actually create jobs. The existing manufacturing workforce is stretched thin, and attitudes toward robots have changed for the better in the last few years – namely, since the COVID-19 pandemic jammed supply chains and spurred a labor shortage.

“One of the barriers to bringing manufacturing back to the U.S. is we don’t have enough skilled workers to bring back as much manufacturing as we could,” Burnstein says. “We’re really going to have to upskill and train the next generation of workers to work in an increasingly automated world.”

Manufacturing’s automation foundation

While cautioning that it’s difficult to predict the biggest advances in medtech automation in the next decade, Burnstein envisions the technology evolving to be able to assist significantly in finding and eradicating cancers and other diseases. Automation and AI have the potential to speed the development time for new treatments, and also to “see” in ways that humans can’t, leading to better diagnostic capabilities.

Burnstein also points out, however, that much of the automation and robotics technology being used in clinical settings today was originally developed for industry. While bin picking and small batch manufacturing may not have the same allure as, say, a lifesaving robotic surgery, these factory floor innovations are foundational to medtech – and without technological advancements in manufacturing, the tools won’t be there to implement new treatments and procedures.

About the author: Clare Scott is associate editor for GIE Media’s Manufacturing Group. She can be reached at cscott@gie.net.

Association for Advancing Automation (A3),
https://www.automate.org

KUKA Robotics Corp.
https://www.kuka.com