June 9, 2022

By Keith Shaw

AutomateTrends FANUC 400x275One of the great things about attending trade shows and conferences in person is being able to see “the big picture” and larger trends within a market based on conversations with several companies and attendees, rather than individual news announcements or meetings. Such was the case for us at this week’s Automate 2022 event in Detroit, where robot manufacturers, component makers and systems integrators gathered to meet with customers and potential customers looking to automate their organizations.

After our meetings and visits with companies on the show floor, we emerged with five key themes and thoughts that show where the robotics space is heading over the next few months and years. These are not presented in any particular order of importance.

Trend 1: Robotic Applications Take Center Stage

Traditional robotic hardware companies are no longer just featuring their robot arm, end effector, gripper, or mobile robot and saying, “Here, come buy this” (although it is cool to see a giant robot arm lift up a car, as pictured above). Instead, they are combining several technologies together and packaging them together as an application solution to solve specific problems.

At this year’s Automate 2022, companies were showcasing how robots can automate the palletizing task, do a better job at welding, and perform automated CNC machine loading and unloading. Traditional gripper makers OnRobot and Robotiq, for example, showcased new palletizing demonstrations and solutions at their booth.

“Integrators are super busy and they will tend to focus on the bigger projects,” said Sam Bouchard, CEO of Robotiq. “The smaller projects, there’s a lot of overhead that doesn’t make sense for them, and it doesn’t make sense for the customers either. So there’s a spot there that we’re trying to cover with small, stand-alone projects that will either be done by the user directly or they will use our solution, which is 80% standard, and the last 20% they will do on their own or with one of our local partners.”

UR CobotWelding800px

At the Universal Robots booth, the leading cobot manufacturer showcased at least 12 different applications that cobots can automate, including TIG (aka Gas Tungsten Arc Welding) and laser welding systems (via partner Cobot Systems), machine tending, dispensing, labeling, screwdriving, 360-degree gluing, inspection and palletizing.

“Everybody wants faster, more precise, and more reach, but you have to temper that with process knowledge,” said Joseph Campbell, senior manager of strategic marketing and applications development at Universal Robots. “If you’re going to build a general-purpose cobot that will be applied in the arc welding space, you have to have an understanding of the requirements of the arc welding process in order to give the OEMs the best possible platform to build on. That’s how we view ourselves – we’re an applications platform, and you see it in just about everything we do.”

Trend 2: Innovations are Making Robots Smarter

For years, large industrial robots were seen as “big, dumb machines” that could do certain things really well and fast, such as lift a car door as part of an assembly process at an automotive factory, or assist with electronics components manufacturing on smartphones. Now, advances in software, machine learning, and vision systems are making these robots much smarter, which opens the door for even better task automation that was previously unheard of.

For example, Mujin was at the Shape Process Automation booth giving a demonstration of a robot de-palletizer, but it wasn’t just a large robot taking the same pile of boxes and putting them in the same space all the time. Through a vision system and Mujin’s robot controller, a pile of boxes could be disrupted and moved into different positions, and the robot would use the vision system to quickly locate the correct place for the gripper to grab a misaligned box and unload it onto a different platform.

SoftRoboticsLotionDeodorant800px

At the Soft Robotics booth, the company was demonstrating a similar task, where its mGripAI system was quickly looking at and picking up plastic “chicken wings” on a conveyor belt that were always in random and dynamic locations. The company also demonstrated an application where a robot and soft gripper could automatically sort two different but similar looking types of bottles (lotion and deodorant, pictured above) onto separate conveyor belts. The combination of software, vision, edge processing allows for this dynamic picking, which could not be accomplished several years ago.

At most of the booths on the show floor, robots were equipped with better software, better vision systems and more intelligence than ever, making them perform cooler tasks than in previous years. This bodes well for the idea that more manual tasks can be automated in the future.

Trend 3: Creating New Processes Based on Data

During at least two of my meetings, I was able to see examples of how mobile robot companies were able to use data from their deployments to help change or improve processes to make things easier for customers.

For example, inVia Robotics was providing demonstrations of its PickWall system (pictured below), which allows workers to fill orders faster while mobile robots can find orders throughout the warehouse and place them on a pick wall without having to wait for the human worker. In earlier versions, robots couldn’t perform their next task until the worker had picked the item that the robot had retrieved, and this could create a backlog or traffic jam if workers at a picking or packing station were distracted or had to go to the bathroom.

“One of the things we know is that people don’t like to be robots, they like to work in bursts,” said Lior Elazary, CEO and founder of inVia Robotics. “We realized that we had to decouple the system to make sure that people can burst, come in, work for a burst and then answer a phone or go to the bathroom and then come back. That’s where we created the PickWall, which is a buffer that works between the robots that can load and unload from a wall, and then the people, who come in and can then pick orders at extremely high rates.”

InViaPickWall800px

At the Zebra Technologies booth, Melonee Wise (vice president of robotics automation, and the former co-founder/CEO of Fetch Robotics) discussed how data from their platform was able to improve how robots interacted with human workers.

“We started working on social navigation behaviors for robots, like if you’re walking down a corridor you walk on the right-hand side, and our robots should move on the right-hand side as well,” said Wise. “What we found is that in casual interactions with the robot where a person interacts with a robot a couple times an hour, by implementing these behaviors we see about a 10% increase in overall efficiency of the robot in terms of velocity and other things. If it’s a high-frequency interaction, like in an e-commerce picking scenario, we saw a 54% increase. It’s these things that not only make it easier for people to work with the robots, but we were also able to pull months of data from the robots before we release the software.”

Trend 4: “One-click” or “One-Touch” Setup of Robot Systems

A trend that became clear at Automate was that in order for more robots to be deployed in warehouses, factories or machine shops, teaching a robot these new tasks will require an easier method than traditional approaches such as using a teach pendant, or highly complex robot programming. Fortunately, many companies were demonstrating “one-click”, “one-touch” or other simpler methods that combined software with vision systems in order to achieve this.

At the Rapid Robotics booth, for example, the company was demonstrating its Smart Setup software as part of its Rapid Machine Operator (RMO) setup to allow manufacturing companies to switch a setup and move between jobs in as little as 60 seconds through built-in vision setups. With a tablet-based app, one button touch has the RMO’s cameras analyze a workspace, and then the system calculates distances and orientations to update motion paths to deliver a fast and precise way to execute a given task.

At the Realtime Robotics booth, they were demonstrating its collision-free autonomous motion planning software (RapidPlan), which crates and autonomously choreographs all robot movements and removes the need for brand-specific robot programming. The company said this could save weeks to months of programming time per project, and also allow for multi-robot interactions within the same cell, without the worry of collisions. 

MicropsiDemo800px

Micopsi Industries was demonstrating an example of how its teach-by-demonstration software could enable precision tasks easier, such as cable plugging (photo, above). At its booth, Micropsi showed a demonstration where a cobot enabled with a vision system and the company’s software was able to plug a cable into an Ethernet port easily without needing to be in the same location every time. The kicker was that the application was “programmed” through demonstration by a “sales engineer in their hotel room,” Micropsi’s managing director, Dominik Bosl, told me.

Trend 5: Time to Deploy is Taking Priority

Most of the companies we spoke with continue to face supply chain challenges to get their systems deployed to customers. Lead times varied from a few weeks to more than 22 weeks in some cases, depending on the complexity of the system and how many different components are needed for the system.

Because customers are facing continual labor shortages, companies that can deploy faster and be more flexible will have advantages over systems integrators or other large companies that might take months to deploy. Companies that discussed different integration projects touted their speed to deploy, which was less prevalent or as high a priority in the pre-COVID world.

The era of customers “tire kicking” products also seems to be over in many cases. “Nobody is buying one AMR anymore, unless they are evaluating the potential for deploying 100 AMRs,” said Soren Nielsen, president of Mobile Industrial Robots (MiR). “With that comes the fleet sizes going up, and the requirements for performance, reliability and services go up, because customers also want us to be responsible for making sure the robots are driving all the time.”

Keith Shaw is the Managing Editor of Robotics-World.com. We plan on producing additional content over the next few weeks from our time at Automate, so stay tuned!