Semi-Autonomous Automation: Right-Sizing for Mid-Tier Manufacturers

Summary

Automation likes repetition. But not all production is repetitive or runs long enough to justify dedicated and comprehensive automation. Many manufacturers face weekly and daily job changes that defeat all but the simplest forms of automation. However, this is changing with evolution of cobots and vision systems that can perform relatively complex tasks in manufacturing facilities with frequently changing production requirements. These technology developments are now enabling mid-tier manufacturers to strike a cost-effective middle ground: semi-autonomous automation.
 

Automation maps to different manufacturing needs

Automation can take any number of forms, but there are three dominant visions of what factory automation looks like in discrete manufacturing.
 
The first, and probably most common, is the fully automated assembly line with interoperating industrial robots that are assembling, welding and painting in one seamless workflow. This automation style is most often found at original equipment manufacturer (OEM) and tier-1 manufacturing operations where production volume and product value are working in favor of full automation. If the job is to build thousands of cars or appliances, then fixed base, industrial automation is necessary for both repeatability and profitability.
 
Probably the oldest form of automation is where custom-made equipment is assembling a few components into a finished product inside the confines of an enclosed, standalone work center. This type of bespoke automation with purpose-built work centers can be found in all tiers of manufacturing operations, but one factor is always present: these are high-volume, long-running jobs. Moreover, the tasks tend to be basic, such as fill, cap and label.
 
The third type of automation is having pick-and-place robots perform simple repetitive operations. It’s certainly common in middle tier manufacturing operations, but it may be more aptly described as assistance versus true automation. Typically, right beside those robots and separated by extensive safety guards, are human operators performing the more complex tasks. The safety factor of moving operators out harm’s way often trumps the cost saving aspects of the automation.
 
Of course, not all products are produced in high enough volumes to financially justify full automation. At the most labor-intensive end of the spectrum are products that are basically fabricated by hand. For instance, very little automation is used in building a one-of-a-kind specialized machine, but hundreds of thousands of products are produced by machines, and then handled, assembled and packaged by human workers. It is this great middle ground of manufacturing that is being most impacted by semi-autonomous automation.
 

Mid-tier manufacturers seek a middle ground in automation

In all of the examples so far—whether it’s building a car, fabricating a work center, or filling a bottle —the components that go into the final product are manufactured by a supply chain. And this supply chain is often comprised of small and medium-sized businesses that specialize in molding, stamping, machining, finishing or assembly and increasingly, some combination of these. Such mid-tier manufacturing companies and their products are the tweeners in the spectrum of automation, facing needs that don’t neatly align with the scenarios outlined above. Their production methods are certainly not hand fabrication, nor are they repetitive or high-value enough to justify full automation.
 
Work centers within mid-tier manufacturers across the supply chain typically include injection molding machines, stamping presses, computer numerical control (CNC) machining centers and assembly benches (as compared to assembly lines). The work center production equipment produces the core part in volume, but the next three or four steps of production are manual. Their batch sizes run in the thousands to tens of thousands, and job changeovers occur daily to weekly. So, they are high-volume enough to look like good automation candidates. But in reality, they are pretty fluid and tricky to automate because the job mix is constantly changing.
 
Historically, the solution has been to simply put people on post primary production tasks, such as inspecting, de-burring, simple assembly, labeling and packaging. For basic manufacturing tasks, the versatility and job changeover time with people is hard to beat—if you have the people, training, work instruction and supervisory team to enable them.
 
However, with recent technology developments, the machine-to-operator equation for mid-tier supply chain manufacturers is evolving.
 

Semi-autonomous automation rises as a mid-tier solution

Increasingly, we are seeing mid-tier manufacturers adopt semi-autonomous automation to reduce their need for operators on jobs that typically run for two or three days and involve molding, picking, deburring, inspecting, lite assembly, packing and labeling. Here, cost-effective automation is now within reach using newer cobots and integrated vision systems.
 
Compared to prior generations of automation technology, the software behind new cobots and vision systems has become so sophisticated that these systems have become an order of magnitude easier to set-up and train and just importantly, move around and retrain. This ease of training means cobots and vision systems can now be cost effectively exchanged between different jobs. They have become in a word: fungible.
 
At the same time, modern vision systems provide a good degree of adaptive correction for imperfectly positioned parts, which reduces the need for workers to monitor production. Plus, cobots and vision systems are sensitive to human intrusion and contact, making safety guard systems either unnecessary or very lightweight.
 
Automating the complete job as described may sound a lot like full automation and in fact, it would be largely automated. But here’s the catch. While running correctly, the work center is automated, but when error conditions occur or a resource becomes unavailable, these types of automation solutions have very little ability to self-correct or self-heal. All they can really do is sense that there is an issue, stop and issue an alert for operator assistance. That is why they are referred to as semi-autonomous work centers.
 
Theoretically, mid-tier manufacturers could invest in either automating work centers around unplanned conditions or creating an operating environment where the chances for unplanned conditions are virtually zero. In reality, the costs and complexity of doing so would be prohibitive. It is kind of like the driverless car problem: it’s one thing to have the computer steer the car down an uncongested highway and quite another to have it self-drive through city traffic in a rainstorm. There are too many contingencies to plan for and environmental factors that can’t be controlled or anticipated.
 
Semi-autonomous work centers that leverage cobots and vision systems provide mid-tier manufacturers with a practical middle ground. They don’t entirely eliminate the need for operators, but they dramatically reduce the number of workers required. Plus, they provide the predictable quality and reliability inherent in automated operations.
 

Conclusion

The automation thought process is evolving from polar choices—either doing the job with operators or making it lights out—to a new middle option that fully automates moderately complex workflows but does not attempt to plan for every contingency or eliminate all possible error conditions. This automation thought process embraces the old saying that perfection is the enemy of the good. By not asking more of the technology than can be cost effectively implemented, it empowers mid-tier manufacturers to take full advantage of the ease-of-use and flexibility of modern robots and vision systems to increase productivity, accelerate growth and improve their bottom line.