Industrial robots are becoming easier to program, more versatile, more cost-effective, more accurate and more mobile. These changes are lowering barriers to entry, shortening return on investment and making robots a more practical investment.
“We have democratized access to robots,” Ready Robotics CEO Benjamin Gibbs said.
Robots, long a staple in automotive manufacturing, also are expanding into other sectors. Collaboration is becoming a key trend—both collaborative robots (cobots) and collaborations between companies to produce improved robotics technology.
“New technology is getting robots into applications that weren’t possible before,” said Bob Doyle, VP of the Association for Advancing Automation (A3). A3 puts on the Automate show every other year.
ARK Invest predicts 3.4 million industrial robots will be sold in 2025, up from about 380,000 in 2017. ARK Investments is admittedly bullish on robot costs, predicting that the price of industrial robots will drop 50-60% by 2025, to less than $11,000 per unit, half the price of other predictions.
“Our goal is: how can we make our manufacturing systems leaner?” Scott Everling, product/business development at Hexagon Integrated Solutions, said. With smaller, more mobile robotic technology, “the cost and inflexibility of installing large machining solutions can be avoided. Manufacturers are looking at these robotic solutions as a way to get manufacturing automation into new areas previously not seen due to the cost.”
Overall, communication between humans and robots has become easier and faster. One example is a collaboration between Rockwell Automation and FANUC America that allows robot- and application-related problems to be identified easier and more specifically.
Although workers still need training, manufacturers have made programming robots easier and more intuitive, Doyle said. Operators don’t necessarily need a four-year engineering or programming degree. Instead high school graduates with a two-year degree and/or an apprenticeship can work with robotic technology and prove to be valuable employees.
Workers no longer need fluency in all the robot languages and develop individual programs for each type of robot in use. Ready Robotics’ Forge Operating System, released in July 2017, runs all of a factory’s robots using the same interface, Gibbs said. With a wink to Star Wars’ C-3PO, the protocol droid that spoke millions of human and robot languages, Forge O/S functions as a sort of translator.
“Before, you had to be skilled in each language,” he said. “Then you had to write ladder logic for the PLC, then write scripts to get the PLC and robot to talk to each other. Then you had to program the tasks. Then you had to debug everything. That could take days or weeks. In one to two days, we have been able to train people who have no robot experience—not only how to use robots but how to reprogram the robot work cell for different tasks.”
Because of improved ease of use, manufacturers who were initially resistant to robots because of concerns about the difficulty of training and the need to bring in specialists to reprogram the system have been able to quickly realize benefits, Gibbs said.
Starting out is “like building a flowchart,” he said. “You chain together building blocks into the flowchart. It’s easy to do a very complex program with our easy interface. You don’t have to use Teach Pendants. You build up the program, drag and drop.”
Individual workers who may have been skeptical about robots are able to quickly come up to speed.
“It’s essentially like using your smart phone,” Gibbs said. “You plug in and we start programming everything through Forge/OS. Language doesn’t matter. Because we’ve made it so easy, people who thought robots were hard have been able pick it up very quickly.”
Customers of Ready Robotics buy the Forge/OS stations and controllers, then pay an ongoing subscription (robot as service) for the operating system, he said.
“Some of our customers have been able to save tens of thousands of dollars in a matter of months, if not weeks,” Gibbs said. “They’ve been able to pay off capital expenditures 50 percent faster. Then it starts to snowball.”
Collaborations between companies also have produced gains.
Rockwell Automation in 2015 released an add-on profile (AOP) for FANUC robots for simpler integration of FANUC robots into Rockwell architecture, said Rob Totten, engineering manager at FANUC.
Prior to this AOP, manufacturers preparing to use FANUC robots had to modify software parameters to match the FANUC robot and run through several configuration screens, Totten said.
“Now, they can pick a FANUC robot from a list of devices within the Rockwell software, add it to the project, add it to your program and drivers,” Totten said. “That makes it a lot easier to add a robot. Once the AOP was released, we got fewer calls on how to deploy robots.”
The automotive sector continues to lead in robot deployments in North America with 52% as of 2018, Doyle said. But that’s a drop from 58% in 2017, as other sectors see more deployments.
Robot deployments are up in packaging with increases especially in food and beverage, according to sales figures from the Packaging Machinery Manufacturers Institute for 2018. Food, beverage and personal care are seeing the most gains with a 13% market share in sales—second only to automotive, according to PMMI research released this year.
One reason for the increased share in food and beverage is new soft robotic grippers, Doyle said. With these grippers, robots can now handle foods such as bread dough and tomatoes. “Before soft grippers, there was no metal type gripper typically used for automotive that could pick up a tomato without turning it into tomato sauce,” he said.
Although still not on par with human performance, robots improved eightfold from 2015 to 2018 in their ability to pick and place items, according to ARK Invest.
In the past, robot measurements have not always been able to achieve needed accuracy for measurements across an entire part, Hexagon’s Everling said. For example, robots could easily perform a task like applying glue near the edge of a structure.
“Looking at the edge of a surface, the robot could apply glue three millimeters from the edge as long as the robot could see the edge,” he said. The challenge came when a robot might need to perform a task, such as drilling holes, across a larger structure.
Hexagon’s Leica Absolute Tracker measuring and inspection system improves manufacturing process accuracy to 50-75 microns, Everling said. Data feedback also improves—to 1,000 pts/sec—enabling real-time control.
A robot manufacturing system combined with Hexagon’s Leica Tracker costs 20% less than technology performing similar tasks, Everling said. “The feedback we get is the whole process is much more flexible, the cost of putting everything together is more affordable.”
But the big advantages are the total size and flexibility.
Hexagon’s Leica Tracker weighs less than 14kg and is 20-30% smaller than other systems, he said. But the real breakthrough is the much smaller size of the total manufacturing system, especially when the robot can be placed on an automated guided vehicle (AGV). The Leica Tracker also easier to pack, unpack and move than other systems.
“Before the Leica Tracker, manufactures used large, custom-made machinery from a variety of OEMs,” Everling said. “Some of these large five-axis machines are 30, 40- and 50-meters long. Once you have one of those in a location, there’s no chance of moving it. More and more companies are trying to get new manufacturing lines up and running in limited space.”
For example, instead of bringing an airplane wing to a robot measuring system, the robot system can move to that wing, he said.
“You can bring it to where you’re already assembling the wing,” he said. “You still have something very large—a robot on a gantry system. But before, you had this huge structure that is built and can’t move. Now, instead of this monstrous tool you’re bringing the wing to, you put the robot on an AGV and bring the AGV to where the wing is. You have this room-sized manufacturing system moving down the wing. It’s a lot more flexible.”
Those improvements open the door for more deployments. Before the Leica Tracker, “the ROI to purchase a very large-scale machine tool for these applications was simply non-existent for some manufacturers,” he said. “We want to make these manufacturing lines more flexible than in the past and remove the infrastructure. That’s the key, the pot of gold at the end of the rainbow.”
Another key step forward in robotics technology came in 2013 when FANUC and Rockwell Automation collaborated to produce enhanced data access (EDA) that enabled more data to be shared between a robot and a PLC, Totten said.
With earlier technology, a robot confronted with a problem sent a basic message saying something like “equipment 1 fault,” he said.
Technicians had to walk—possibly 100 yards and up or down a flight of stairs—to the robot teach pendant to identify the nature of the fault, address the issue, then walk back to the HMI (human machine interface) to restore the system to automatic.
“Depending on the layout of the line, that’s a lot of walking,” Totten said. With EDA, the robot or the PLC integrated with the robot can read and describe the specific fault, such as waterflow fault or weld controller fault, in a display at the HMI or on the production bingo board, he said. Then the right technician can go directly to the robot or other equipment to fix the problem.“The real time savings is the access to the equipment and not necessarily having to cross a line to get to the robot,” Totten said.
EDA also allows for easy tracking of trends, he said.
“Knowing that ‘robot 3’ at an automotive plant faulted 10 times over the course of the last shift is one thing,” he said. “Better to know all 10 faults were in this particular job or this style of vehicle or these two process locations” because that kind of information can be used to help schedule maintenance, as well as to review specific programs or applications.Emerging technology like collaborative robots (cobots) and AGVs and autonomous mobile robots (AMRs) are drawing attention to robotics, although companies may eventually decide that a traditional robot makes more sense.
Cobots were introduced at automation trade shows around 2013, and by 2019 many trade show booths had some kind of collaborative application, Doyle said. Interest grew at a similar pace for AMRs, which entered the trade show scene around 2015.
“Certainly, collaborative robots are designed to have simpler human machine interfaces that allow people to program the robot more easily,” he said. “Some companies have even offered hand-guiding programs and more user-friendly teach pendants.”
Cobots and AMRs are not always the answer. Both pose challenges.
“A potential end user might read about a mobile robot or see something about Amazon robotics and think mobile robots solve all distribution challenges,” Doyle said. “Users have to determine: Does an AMR really bring in new opportunities?”
For example, although AMRs eliminate the need for the magnetic line on the floor needed for AGVs, AMRs require reliable WiFi, which isn’t always available in a factory, he said.
And, “collaborative robots are typically slower and have lower payloads because they are power-force limited and can’t run at high speed in case there is an opportunity to hit someone or something,” Doyle said.