Robotics and automation have never been hotter. And it’s not hard to see why. Squeezed by the ongoing e-commerce boom and a nationwide labor crunch, companies are finding they need these technologies to boost fulfillment speed and accuracy at a time when adding people is no longer an option.
To learn more about the current state of the robotics market as well as what lies ahead, DC Velocity Group Editorial Director David Maloney gathered experts from companies that participated in DC Velocity’s Robotics Forum at last month’s Modex show for a freewheeling discussion. Among the topics we address in Part One (Part Two will appear in our May issue) are recent advances in robotics technology, the different ways in which robots can help ease the DC labor crunch, and the future of robot design.
Jeff Christensen, Vice President of Product, Seegrid
Dan Coote, Global Product Manager, Locus Robotics
Mike Futch, President and CEO, Tompkins Robotics
Matt Kohler, Director of Applications, Bastian Solutions
Romain Moulin, CEO, Exotec
Divya Prakash, Director, Business Consulting Industry 4.0, SICK
Kevin Reader, Vice President of Marketing, Knapp North America
Nicola Tomatis, CEO, BlueBotics
Q: What do you consider to be the most significant advances in robotics technology within the past five years?
Jeff Christensen – Seegrid: The most important advancements today are in the development of intelligent autonomy technology. In order for autonomous mobile robots (AMRs) and automation solutions to gain an extremely accurate understanding of their immediate surroundings, they need to see, collect, prioritize, and interpret a higher density of real-world live information. Any single sensor when used alone has its limitations. However, robotics companies that embrace sensor fusion will have solutions that offer an enhanced ability to perceive, plan, and control movement.
Matt Kohler – Bastian Solutions: Artificial intelligence (AI)-powered vision is a game-changer for robotics. Robots have always been great at handling repeatable tasks, which has led to their widespread adoption in manufacturing, but not so much in distribution, primarily due to the wide variety of products you see in a distribution center. AI-powered vision helps fleets of robots learn how to handle the complexities involved in handling tens of thousands of SKUs and the random nature in which they are presented to robots.
Nicola Tomatis – BlueBotics: One of the biggest changes in the last five years is that automated guided vehicles have become both the fastest-growing market and the largest market in service robotics. AGVs have overtaken both industrial robots and service robots working outside of industry.
Mike Futch – Tompkins Robotics: Vision systems have made significant strides. Many picking software providers have made continual improvements in these systems. Two of the biggest hurdles that picking vision must overcome are the wide variety of items in warehouses and the unpredictability of what items are in a picking container. While human intervention is still required, picking vision has progressed to the point where the systems can economically replace or supplement manual labor in the warehouse environment.
Kevin Reader – Knapp: Developments in artificial intelligence and machine learning have rapidly enhanced the capabilities of the latest generation of more capable robots, enabling them to perform more complex order picking tasks. At a time when labor shortages have risen to an unprecedented level of concern to supply chain practitioners, this category of robotics is moving from prototype consideration to the “proven business case” category.
Divya Prakash – SICK: Simplification of user interfaces has opened up the robotics technology market to a much wider audience of end-users. Also, increased processing power enabled by GPUs (graphics processing units) and stronger CPUs (central processing units) better supports enhanced sensor capabilities.
Dan Coote – Locus Robotics: The most important advancement in robotics technology lies in how robots are able to navigate around a warehouse. The granularity of a map within a warehouse along with the bot’s ability to navigate in a complex and convoluted environment both safely and efficiently have progressed by leaps and bounds. It’s a combination of the physical hardware that goes into the bots (including LiDAR sensors) and enhancements in the back-end software tools.
Romain Moulin – Exotec: Battery technologies have improved a lot in terms of efficiency, power density, and fast charging, which has allowed for new applications in mobile robotics.
Q: How can the increased use of robotics help distribution facilities cope with labor shortages?
Matt Kohler – Bastian Solutions: We hear about labor challenges all the time, and of course, implementing robotics in a distribution facility will help reduce labor requirements. But it also allows companies to re-allocate that labor to more value-add areas of their operations. I believe that is important because it will help those employees become more engaged and [will] hopefully increase retention.
Mike Futch – Tompkins Robotics: Robots have a bigger impact than just labor replacement. Consider that today’s workforce seeks out technology. They would much prefer to work in an environment where they can interact with high-tech solutions over traditional manual processes or even traditional automation. Further, facilities are continuing to see positive impact by adding in “gamification” to their operations. Robotic solutions greatly enhance this ability.
Dan Coote – Locus Robotics: Robotics reduces the requirements for the number of warehouse workers and increases productivity among the workers that a warehouse does have. During peak season, there’s enormous demand for people, and the number of people available is dependent on where they’re located. To counteract this, bots—and specifically RaaS (robots as a service) bots—offer the option of short-term rentals, which cuts down on the number of people required and provides the ability to [meet] your peak demands by augmenting throughput with a combination of people and robots. Also, the induction, training, and speed-to-competency times decrease with robotics.
Q: How will artificial intelligence and machine learning affect future robotic designs?
Kevin Reader – Knapp: AI and machine learning solutions have the capability to impact a great many areas, from manufacturing through operations, delivery to customers, and beyond to our daily lives. But in a general context, there is consensus that the future of robotic design is not just driven by AI, but that “cloud is the enabler, data is the driver, and AI is the differentiator.”
Matt Kohler – Bastian Solutions: The biggest impact will be in the sheer number of different robotic applications that those technologies enable. But from a robotic system design standpoint, I see the biggest impact being on “end of arm tooling” (EoAT) designs. AI will be able to evaluate products presented to the robot and determine how best to handle the product. AI can tell the robot to use only certain suction cups on a particular product surface or to only clamp a product in a certain location, or it may even tell the robot to swap out its current EoAT to use a different style of EoAT to handle a particular product.
Jeff Christensen – Seegrid: Robot design will benefit from AI and machine learning as the software will increasingly take on decision-making. Industrial automation solutions will move from robots that have to do many things to robots that will be more task-specific. This specialization, along with coordinated flow and interoperability across these disparate robot types, will enable process innovation and ongoing AI-driven optimization. Not only will movement be automated through the robots, but the optimization of flow and continual improvement will also be automated.
Divya Prakash – SICK: As artificial intelligence and machine learning advance, they will allow robots to complete more complex tasks in a shorter time. In the context of increased sensor capabilities, such as being able to localize mixed parts in a bin, this will lead to more advanced robot capabilities to support the additional sensor capabilities. Seven axis arms may become more common. More advanced flexible tooling (soft grippers) will come to market. Ease of use will need to be maintained or improved.
Romain Moulin – Exotec: Artificial intelligence, coupled with new generations of sensors, allows robotics to tackle operations that once could only be done by human operators. Bin picking is still the most typical application.
Q: How has warehousing software evolved to make it easier to integrate the robots that are taking on a growing role in fulfillment operations?
Jeff Christensen – Seegrid: In the past, warehouse software interacted with automation such as AMRs and automated storage and retrieval systems (AS/RS), but had no visibility to the progress of work once it entered these systems. Tighter levels of integration through generic APIs (application programming interfaces) give today’s warehouse software much better visibility into automation and robotic technologies. This also allows for changes to either system without time-consuming and costly integration changes.
Kevin Reader – Knapp: As with most technologies these days, when it comes to warehouse and fulfillment operations, software is not stand-alone but rather, integrated across an array of areas—from overall inventory management to sequencing, order picking, device control, and daily operation. One of the most significant improvements is the development of control tower technologies that focus on warehouse and fulfillment center management and are designed to manage overall resources. Best-of-breed warehouse management systems are also being re-architected for the real-time nature of the latest technology, including goods-to-person systems, robotic palletizers, and fully robotic order picking.
Nicola Tomatis – BlueBotics: Software has improved in two key areas: human usability and machine interfacing. Warehouse software today is easier to use, while at the same time, it needs to interface with more vehicles and systems than ever before. It has improved, but a lot remains to be done.
Q: What will robotics look like by the end of the decade?
Nicola Tomatis – BlueBotics: Two trends will define this decade in robotics: ease of use and interoperability. Robotic systems will be simpler to put in place and easier to use on a day-to-day basis. We will also see more and more different types of robots—and even brands of robots—working together as interoperability becomes the default.
Romain Moulin – Exotec: More and more operations associated with picking, packing, and loading/unloading will be fully automated.
Divya Prakash – SICK: The merging of sensors and robotic systems into a more cohesive, pre-engineered application-based product will allow for more advanced applications to be solved. Rather than customizing automation to existing processes, existing processes must mold to the needs of the automation system. Some of the newer tasks being attempted with automation are complex enough that there may only be one or two ways to solve them. This contrasts with the past in robotics, where there were many ways to solve an application and customize to the user’s existing processes.
Dan Coote – Locus Robotics: By the end of the decade, it will be commonplace to have a multitude of bots performing various functions across the warehouse or [robots from] a multitude of vendors deployed within the warehouse.