Most people are aware of the newest developments and the latest generation of robots and artificial intelligence (AI). Much has been made of how these things will bring changes to manufacturing and will impact jobs. Medical device manufacturing can clearly benefit as much as any industry with these technologies. But our industry isn't selling entertainment. We're focused on safely and reliably manufacturing products that improve and save lives. Poorly implemented robotics or AI can have serious consequences well beyond simple yield issues.
Equally important, if the next generation of people who might fill the upcoming manufacturing positions come to believe the hype about AI, we will lose the pool of talent that powers our manufacturing businesses. So, it is important to separate fact from fiction to determine what is a real value proposition and what is not true technically and career-wise in the future.
Most are aware that we are already on a path leading toward significant changes in how we manufacture products. Historically, manufacturers across the world have relentlessly chased the cheapest labor for manufacturing products. This has led to intense use of automation. As the products being manufactured evolve, so do the demands on the processes used. Until recently, the use of robots as labor-saving devices required only basic pick-and-place operations. These are essentially very simple operations with machines that are equivalent to a person wearing a blindfold and using heavy gloves.
The advent of cheaper cameras and sensors has kicked off a revolution in robots. Add to that cheaper, faster computer processing horsepower, and the stage is set for AI to bloom and quickly turn pick-and-place robots into an old hat technology. This revolution in technology will enable less-expensive robots to perform assembly, something that has so far been reserved for more expensive robots and fit-for-purpose mechanisms.
As with any change, there is a combination of great opportunity and great trepidation. Much is made of the potential loss in jobs as a cost of this advancement in automation. Yet history suggests that changes in technology will also change the opportunities available for those who have the courage to change and adapt to the new opportunities.
By example, one might remember a bit of history about accounting. At one time, there were rooms filled with people operating hand-cranked adding machines. They totaled numbers, checked each other's work, and rechecked each total. Each person pulled the handle of their adding machine to enable a mechanical calculation of numbers.
Electronic calculators, main frame computers, and ultimately the PC changed all that. There are no more mechanical hand-cranked calculators. But one would ask, are there fewer accountants? Clearly not. But they do different things now. Things of higher context. Things the machines can not do. A host of new jobs were created, but they required people to go back to school and retrain. Some adapted, some didn't. Yet this new technology helped create at least two new industries at the time: computer hardware and software. Two small industries that didn't create many jobs, right?
Robots and AI bring the promise of new industries and new jobs, but clearly change is required. Computers and automation have already brought change, but the question is: what is different about these new changes? It is important that we help the upcoming generation of manufacturing workers understand that there are career opportunities.
Robots vs. Mechanisms
Many people confuse some basic mechanisms with more complex ones and call machines robots while they are nothing more than high-tech CNC machines. Such machines follow very rigid “if-then-else” structures. They handle important process variations quite badly. Such machines, no matter how many motors and gears they have, do not fit the definition of an AI-equipped robot. For a clear definition of a robot with AI for manufacturing, one can fall back on something akin to Isaac Asimov's vision of robots. For this purpose, this will be our definition of an AI-equipped robot:
An Asimov Robot performs a set of tasks in a goal-oriented manner. This behavior is demonstrated by successfully handling significant changes and variations in the operating environment that force differing solution paths. Anything else is just a mechanism.
Why is this important? Those of us who develop and deploy automation know something that the general public consistently misunderstands. People completely underestimate the complexities of their own hand operations. And they often overlook the complexity and sophistication of their senses and how those senses are used to guide and control their hand operations. And even more importantly, most people seriously underestimate how massively complex those things are when combined with dynamic goal-oriented behavior.
There is no affordable robot today that can match even just a visceral level of intelligence. The given definition for an Asimov Robot embodies a requirement to have at least this level of capability. For example, have you ever seen a robot that can assemble an electronics harness, test it, pick up a PCB, and gently plug that harness into place? People underestimate the coordination and skill set required just for that simple operation. Yet that is the level of intelligence we need to see in robots in order to really step up manufacturing and bring it home.
The reality, however, is that just adding vision was a recent big addition to robots, and we still do not have anything close to Asimov Robots in common use. Only companies with deep pockets can afford the investment required to make advanced robotics a good value proposition. For robots to become practical and ubiquitous to manufacturing, a combination of business and technical requirements must be met:
Affordable total cost and payback period — this includes the costs to install, set up, build special fixturing, create special tooling, and program or hire people to program the robot, as well as operation and maintenance costs. Having a robot that costs less per year than the cost of a person is not enough of a value proposition to invest if it entails a substantial payback period. A long payback period can wipe out any advantage in low initial cost.
Enhanced adaptability — many manufacturers utilize high-mix product lines. Multiple products are often manufactured in the same floor space by performing quick line changeovers. This is the impetus for developing general-purpose robots that can be easily adapted to new tasks on a regular basis.
Fast deployment — from the moment the product development is far enough along to kick off the creation of a manufacturing line, how long does it take to have a robotic solution that fits with the manufacturing plan? What is the lead time of special fixtures, tooling, programming, and process verification and validation?
Robots for Medical Manufacturing
When evaluating any robotic solution for medical device manufacturing, there are some additional aspects that can factor greatly into any decision to deploy them into production, specifically in the area of process validation:
If the AI used in a robot is based on any type of statistical, nonlinear, or in-process learning algorithm for how it will execute the tasks assigned, how do you prove that every solution path will meet product requirements?
If history is any indicator, AI software will require constant updates. How do you revalidate each update of the robot basic operating system in your application?
These two factors alone could spur a new career in manufacturing. Someone must become expert at answering such questions and providing application-specific solutions. These are perhaps not quite up to par with Asimov's concept of a robopsychologist but, nonetheless, they crate a new people-based career specializing in robot applications.
But one potential advantage of Asimov Robot's becoming commonplace is to help open the market where small-to-medium sized businesses (SMBs) can capture and hold manufacturing positions in the face of larger, global competition. Why would this be important to SMB medical manufacturing? Of the many things these robots will enable, this technology can spark device innovation and manufacturing in ways not open to SMBs in the past.
When the cost of assembly labor falls, and when outsourcing that labor is no longer required, design and manufacturing engineers will have more options when developing products for SMBs. Complex assembly processes that used to be out of reach of SMBs can now be brought within their grasp economically.
There is an analogy in software here. When the cost of computer memory fell dramatically, one no longer needed to be an expert in absolutely minimal size programming. This meant that applications and products could be created faster. This, in turn, led to the development of operating systems that hogged memory but allowed an entire group of people to write code without having to know anything about the hardware it ran on.
The argument is that making Asimov Robots an economical tool available to SMB medical manufacturers has the potential to let product developers and manufacturers focus more on innovation and the efficacy of their products. Less time needs to be spent on designing a harness to make it easy to manufacture; less time and less cost of working to optimize the number of screws used in an assembly; more time getting the overall product right and to market. These new tools can open the door to using processes previously deemed too expensive, out of reach of SMBs, and available only to large-scale manufacturers.
If Asimov Robots become commonplace, it can become economically viable for SMBs to be built around families of low- to medium-volume products. This potentially opens the door for a broader range of medical devices that are affordable while providing adequate gross margins to support sustainable businesses. This could enhance innovation and provide a greater choice and variety of medical solutions in the market.
It has been shown many times that the innovation engine of industry has a strong home in SMBs. That innovation came from people who envisioned new opportunities where there were none. People who had the courage, took risks, embarked on learning new things, and had the perseverance to work through times of change. People who saw the need to change to make a better future. Just remember, robots have none of that. But people do.
This article was written by Robert L. Kay, President and CEO of Elite Robotics Corp. (Camarillo, CA). For more information, Click Here.