The Wrong Way to Automate and How to Fix it

Faster work cells aren’t always better for small businesses, but this Montana-based manufacturer found the sweet spot

Universal Robots’ UR10e collaborative robot at work in the MT Solar facility. (Image: Universal Robots)

Universal Robots’ UR10e collaborative robot at work in the MT Solar facility. (Image: Universal Robots)

Many small manufacturers are investigating automating their operations: not only is the technology more affordable, it’s also easier to set up, manage and integrate with a company’s other processes than ever before.

MT Solar is a Montana company that makes stands and fixtures used in solar panel installations. The firm recently incorporated automation into its operations to meet a 300 per cent increase in demand for its parts every summer; the company has experienced challenges in finding skilled workers to help meet the demand.

MT Solar focuses on fast lead times for making its products. The company had staggered sales and low volume. “The dollar value tied up in inventory for a fast delivery solution allowed us to help our customers feel good about buying from us versus the competition,” said Travis Jordan, founder of MT Solar. “But about four years ago, we had hit the breaking point of that process and we had maxed it out. Our batch sizes had been brought down as small as we could bring them.” The company brought in a CNC plasma cutter—it helped speed up the production of parts, it wasn’t helping the company meet its end goals. It had plenty of inventory of finished products, but at a level that wasn’t ideal for business.

The Wrong Way to Automate

“It was a classic case of the wrong way to automate,” said Jordan. “It’s the right way to automate for maybe a car manufacturer, but we’re buying it from our local welding shop. So we have a disjoint between robot welding that that requires high speed and a shop who’s got a batch run of 500 parts and wants to start to automate.”

Jordan compares it to a grandfather building dining tables for his children and grandchildren. “If he’s going to build one table, he might as well build four because takes very little extra time to build four than it does to build one,” he said. “Problem is, if he’d had 16 grandkids, he’d have been stumbling over piles of table legs to try to get things done.” That’s why scaling operations for a small manufacturer, particularly one that works in batch orders, may not be the right approach.

MT Solar finally landed on a cobot as the solution. “We said, what do we need to deliver out the door to the customer? And how many of those do we need to deliver per day?” said Jordan. With those targets in mind, the company created an assembly line that combined manual processes and the cobot. “Maybe you’re at a rate of 10 or 12 per day,” said Jordan. “You divide the day up into blocks, then you divide all your segments of all your parts into blocks and you say, how many of these parts can I produce in that time block? How many of these parts can I use the robot to produce? Well now I have 45 minutes for this robot to work, which is a long time in an automation space.”

Under this approach, it’s not important how fast a robot can make the part, but rather how repeatable and how well can it make the part without human intervention. “I don’t care if it takes it 20 minutes to make the part and the robot can do it in five,” said Jordan. “If I only need one every 20 minutes, this robot’s the right answer for me.”

MT Solar acquired the cobot from Vectis Automation, an automation integrator—which is a company that installs and starts up technology in a manufacturing setting.

“[MT Solar’s] peril was similar to the majority of manufacturers today—it’s tough to find skilled-trades labor, and those skilled workers don’t want to do the boring work anyways (nor should they),” said Josh Pawley, Vice President of Business Development at Vectis. “The business aspects of increased productivity, a more inviting place to work, and improved quality and lead times were all factors too. They were able to easily offload the ‘boring’ weldments to the cobot and increase productivity on them at the same time.”

Jordan found that the cobot not only improved the manufacturing process, it also helped his employees improve their skillsets and perform better. “The cobot allows me to get collaborative with my 16-year-old kid who can run the little tablet and my 60-year-old gray-haired guy who’s been welding all his life and says, that just doesn’t sound right.” MT Solar’s employees don’t have to be technologically savvy to work with the cobot. “There’s no real entry barrier there,” said Jordan. “I’m not scared to throw somebody new at it: they pick up the basics in 15 minutes, and then you can have somebody tweaking stuff.”

Keep the End Goal in Mind

Through MT Solar’s experience in getting into automation, Jordan concludes that there are two paradigm shifts that have to occur for small companies to become efficient at automation. 

“The first paradigm shift is you have to accept that you’re going to solve the small problems first,” he said. Many companies, understandably, may want to focus on automating the most complicated problem first—the one that gives them the most trouble. Jordan thinks that’s the wrong approach. Instead, companies should consider finding efficiencies and solving smaller problems first, to optimize their existing processes as much as possible before automating them. “Maybe your first step of automation might be moving your pallet one step closer to your jig so that you can take one step rather than two to get something set,” he said. “You save a minute here, a minute there, a minute on the next one.” Jordan tends to hear lots of pushback on that approach from operators, who don’t think it’s worth automating something that only saves them a minute. “Automate the easy things first, grab your low-hanging fruit, because then that buys you time and money to begin to tackle the more and more difficult ones,” he said.

“The second biggest key point to automating is to make sure you get your end goal of your measurable output solved first and then focus all of your automation against the end goal,” added Jordan. Manufacturers may take the approach of wanting to automate each piece of the process, assuming the sum of the efficiencies gained at each step will result in time saved in making the finished product. However, changes to each step need to be oriented to the broader goal of efficiency—not just making each stage faster.

Even if it takes longer to make each individual part, time can still be saved in getting finished products out the door in time if the process is automated the right way,” said Jordan. “Why not cut 30 beams at once? The saw is going anyway—cut the whole bundle at once. But what do I do with the 29 beams I didn’t need? I have to stack them on a pallet, and later on I have to reach down, pick them up and reposition them. That time savings of cutting that many parts I didn’t need was burning up my guys because they automated the work cell rather than automating the process.”

Finding the right automation technology for the job, and not automating for the sake of automation, helped MT Solar respond to demands from its customers and improved the company’s manufacturing processes—demonstrating how small manufacturers can make use of automation to maximize their busines