Issue link: https://iconnect007.uberflip.com/i/1126810
56 DESIGN007 MAGAZINE I JUNE 2019 onous, dangerous jobs—and ideal for automa- tion. In the debate over worker displacement, people seldom mention how automation has improved worker well-being. Robots perform numerous jobs that otherwise pose health and safety risks, whether it's a major risk, such as toxic fumes, or a minor one, such as repetitive muscle strain [2] . In these and virtually all prac- tical applications of robotics, a human being is still monitoring, controlling, or programming the machine. In addition to minimizing injuries and deaths, robots excel at routine tasks that re- quire a high degree of accuracy, repeatability, and consistency. In PCB manufacturing, such tasks include the assembly of service mount components as well as picking and placing. From chip shooters to gantry-style machines, our industry has been using robots for decades. Can You Teach an Old Robot New Tricks? No matter how fast or advanced a facility's robotics, humans need to direct, troubleshoot, and keep track of the work. Moreover, outside of high-volume production runs, PCB manu- facturing is not an assembly line. It's hard to imagine how an automated system could be useful in low-quantity prototyping or any other non-linear, one-off process that demands mul- tiple judgment calls. That includes even seemingly simple pro- cesses. Say you needed three sheets of five-mil laminate with two-ounce copper thickness. Now, consider how many steps are involved in identifying, fetching, and stacking the ma- terials. You could feed instructions into a ro- bot, hit "go," and the robot will do the job ac- curately, but will you save any time? Probably not. Nor are you likely to save time in the long run. At Sunstone, each order is unique. With ro- bots, we would need to program every step in the manufacturing process for every order. It's much faster and easier to fabricate by hand. On top of that, unlike the computers we buy, humans don't become obsolete—or at least our obsolescence takes longer. Employees can learn and be retrained. Machines need to be updated or replaced. Artificial Intelligence vs. Artifice "But wait," you might be thinking. "Aren't machines already taking over complex human jobs? What about artificial intelligence?" If you think AI is close to approaching anything re- sembling human cognition, here's some bad (or possibly good) news: What we call "AI" is hardly ever intelligent. According to an article by James Vincent ti- tled "The State of AI in 2019," true artificial intelligence is a long way away, and genuine breakthroughs tend to get buried under mar- keting "hype and bluster" [3] . The article names Oral-B's Genius X toothbrush, which debuted at this year's Consumer Electronics Show (CES), as one example of the phenomenon. The device's "supposed 'AI' abilities" are really based on "clever sensors" that provide "simple feedback about whether you're brushing your teeth for the right amount of time and in the right places." Dentists can rest easy, for now. How "Futuristic" Will the Future Be? Make no mistake—plenty of exciting devel- opments are happening in the world of robot- ics; they just aren't the kinds that will lead to a Terminator-esque dystopia. Boston Dynamics, for instance, has built machines such as BigDog and Atlas that dis- play impressive, almost lifelike abilities to ma- neuver around simulated battlefield environ- ments. Boston Dynamics' creations are not autonomous, however—a Wired interview with the team reveals movement is "mostly done by humans and remote control" and the machines "follow a well-defined set of rou - From chip shooters to gantry-style machines, our industry has been using robots for decades.