Image: MiR

Autonomous mobile robots (AMRs) are one of the most exciting technologies in the robotics and automation sector. AMR deployments for warehouse and industrial operations are surging, and this growth is expected to increase for the foreseeable future. According to a market report published by Next Move Strategy Consulting, the global AMR market generated $1.61 billion in 2021, and is estimated to generate $22.15 billion by 2030, witnessing a CAGR of 34.3 percent from 2022 to 2030.

The lack of an available workforce willing to take on non-value-added roles is a key driver of AMR adoption. By taking on many repetitive tasks in the warehouse, AMRs can deliver everything from order picking to moving pallets and large payloads. As the technology advances, engineers are implementing new software capabilities for AMRs to successfully navigate different environments.

How important is the integration between the hardware and software? How will AI influence the next generation of AMRs? Tech Briefs posed these questions to five industry experts to garner their views on the current status and future outlook for AMRs.

Kevin Dumas, MiR

Our roundtable participants include Kevin Dumas; VP of Products, Mobile Industrial Robots (MiR); Steve Branch, VP of Sales Engineering, Locus Robotics; Jay Judkowitz, VP of Product, OTTO Motors; Ed Volcic, VP of Industry Management, KUKA Robotics, North America, and Joshua Alphonse, Head of Mobile Robotics, ABB, U.S.

Tech Briefs: AGVs have been in the market for a long time. How are AMRs different and changing the impact of automation?

Kevin Dumas: AGV technology traditionally required some sort of physical guidance embedded in the floor, like tape strips or magnets, that the AGV always follows. AMRs are a smart technology that has a “planned” route, but that route can be changed by “re-mapping” and changing the assignment. AMRs are also adaptable or flexible, as they can recognize unplanned or unanticipated obstacles and navigate around them. This is something AGVs cannot do. Typical tasks for AMRs include tugging racks of material in a tow-behind approach, picking racks of materials or products, restocking of critical production operations, and then simple tasks such as trash removal.

Steve Branch, Locus Robotics

Steve Branch: While AGVs have been used for many years, they typically follow pre-determined routes and lack the flexibility — and autonomy — that AMRs offer. Long, pre-determined AGV routes lead to inefficiencies and poor asset utilization. For example, if an AGV encounters an obstacle, it typically waits until the obstacle is cleared. AMRs, on the other hand, autonomously navigate around obstacles or plan alternate paths to the destination. AMRs are more flexible and adaptable — and safer — to navigate dynamically changing environments independently and safely work alongside humans.

Jay Judkowitz, OTTO Motors

Jay Judkowitz: A lot of our peers like to market AGVs vs. AMRs, but OTTO Motors sees it more as a continuum. Historically, AGVs were very good at following physical guides and staying on a perfect path, much like a train stays on the tracks. If you look at an AGV deployment, you can see ruts in the concrete from the repeated driving on the exact same path. AGVs are predictable and not smart individually. The intelligence to drive at scale had to come from the fleet manager which works like an air traffic controller, telling the robots exactly what path to take, when to slow down and yield, etc. This works really well in a predictable environment that is homogeneous without pedestrians or manually driven traffic doing work in the same space.

AMRs have taken the other extreme. Like offroading ATVs, they can go wherever, whenever. They have an endpoint in mind and will get there any way they can. They can take any path they fit in, no matter how surprising. They won’t quit. They can avoid local obstacles and replan globally if completely blocked. It’s great for deployments when the job definitions or facility layouts can change. Because of the lack of physical guides and the robots’ ability to figure out paths on their own, reprogramming is trivial. But the unpredictability can lead to behavior that surprises bystanders and the fact that every robot makes its own decisions without an air traffic controller equivalent can make traffic backups a problem.

But, like I said, there is a convergence happening. AGVs are doing more “virtual line following” and eliminating the need for physical infrastructure. Some AGVs will do local obstacle avoidance. Some AMRs do virtual line following rather than ATV-like free form roaming. Also, the gap between the two is closing in the management and standards space. Historical AGV standards like VDA5050 are starting to include AMR concepts so that there can eventually be capable unified management across the two mobile robotic technologies.

Image: Otto Motors
Ed Volcic, KUKA Robotics

Ed Volcic: AGVs have been in the market for a long time and are typically used for repetitive tasks such as transporting materials from one location to another using methods like tape or magnetic guidance to follow specific paths. AMRs are different in that they are more flexible and can adapt to changes in their environment without requiring significant modifications to their programming or infrastructure. AMRs are typically equipped with advanced sensors, such as laser scanners and LiDAR, and algorithms that allow them to navigate complex environments and avoid obstacles in real-time.

The impact of AMRs on automation is significant as they have the potential to transform the way that materials are transported and processed. With their ability to operate autonomously and collaborate with humans, AMRs can also improve productivity and safety in the workplace. AMRs can also be paired with one or more articulated robots on top of them. This allows a single articulated arm to tend to different stations, e.g., CNC machines, 3D printers, semiconductor automated material transport and machine tending etc., thus significantly reducing costs.

Joseph Alphonse: AGVs have been on the market and implemented in manufacturing since the early 1970s. These systems involve a vehicle which is set on a fixed track to conduct logistic missions. This track is typically a magnetic strip or QR codes which define a fixed path. When the vehicle is presented with an obstacle, the vehicle must wait until the path is clear before it can resume operation. AMRs are not fixed to a set path and have the autonomy to navigate around obstacles. The technology in AMRs allow for a much more flexible and adaptive environment where pickup and drop off locations can be quickly modified, and vehicles are much more efficient in executing their missions by continuing their travel regardless of changes in the work environment.

Tech Briefs: What are some of the new technologies being introduced to facilitate easy implementation and safety of AMRs? How will AI influence the next generation of AMRs?

Kevin Dumas: 5G is one new technology that will facilitate AMRs. Today’s mobile robots don’t require much infrastructure, so companies can easily start with one robot. However, as soon as facilities add additional robots, they must have a way to connect and manage them. For now, at least, that depends on Wi-Fi. Unfortunately, most Wi-Fi infrastructures were not built for sharing data at this level. Once 5G wireless networks are widely deployed, its stronger and more robust IT infrastructure will be a game changer for AMRs, enabling companies to efficiently deploy and manage dozens or even hundreds of robots.

Software is also critical to AMRs. For individual robots, improvements in sensor technology and connectivity advances such as 5G will help feed data into AMRs’ planning algorithms and software. This data is critical when navigating around facilities, enabling the robots to see any object around it or from above it. Until recently, even two or three AMRs were considered an advanced set up. With facilities now deploying large numbers of AMRs, there’s a greater need for coordinating, controlling, and managing the fleet with a single internal logistics system. From a web-based interface, fleet management software improves the efficiency of AMR operations, enabling easy programming and control of the robots, including those with different top modules, hooks, or other accessories.

As for AI, although it will be some time before robots can extrapolate outside of training, a new generation of algorithms allow robots to be more effective at learning from what they experience to better detect and make intelligent choices in new situations. Today, advanced AMRs rely on AI for capabilities such as identifying objects with greater precision and informing navigation for safe maneuvering through busy facilities, all with less power consumption. And while huge jumps in AI capabilities are still to come, the autonomous behavior of AMRs — how they get from one point to another without truly knowing what goes on around them — will continue to improve.

Steve Branch: Technologies such as advanced sensors, machine learning algorithms, and cloud computing are already being used to enhance the implementation and safety of AMRs. Integrating AI is pivotal in the evolution of AMRs, empowering them with capabilities like predictive analytics, advanced decision-making, and adaptive learning, helping to make them even more efficient, autonomous, and safe to work alongside humans. Sensors such as LiDAR, used in many AMR solutions, continue to improve and are providing longer-range and more precise visibility of the environment. AMRs can now perform more advanced decision making, improved and earlier obstacle avoidance, and generally operate more efficiently and human-like. Also, continued advancements in SLAM (Simultaneous Localization and Mapping) technology can now automatically update maps and capture changes in the warehouse environment. Because warehouse environments are very dynamic, this helps to provide continuous AMR operation with minimal to no downtime when changes occur.

Jay Judkowitz: From a usability perspective and ease of implementation, OTTO Motors feels very comfortable that we’ve introduced the technology needed to make the robots as easy to implement as possible on their own. However, the robots don’t work in a vacuum. Often their jobs are called from higher level ERP, WMS, or MES systems. The robots and their attachments need to be connected to factory endpoints. Both these points of integration require programming, either using REST, OPC-UA, or PLC programming. That programming is different for each robot vendor. This integration will get easier over time. I think that the ERP, WMS, and MES vendors will need to work with AMRs out of the box so that the integration does not need to be reimplemented by each customer. Third party software providers are doing good work to streamline endpoint and attachment integration. Third party software vendors are also starting to work on single panes of glass for managing multiple robot vendors. And, that effort will be made easier by expanding standards and open source.

Image: KUKA Robotics

The potential use of AI for this space for driving intelligence, fleet optimization, and customer education is limitless. That said, the clearest near-term use of AI is in computer vision. That’s a well understood space and if your robots can understand what they are seeing around them, they can make much better driving decisions to limit slowdown and congestion, decrease time, and improve ROI.

Ed Volcic: Several new technologies are being introduced to facilitate easy implementation and safety of AMRs. One technology is LiDAR sensors that use lasers to create a 3D map of the environment and can detect objects in real-time, allowing the AMR to navigate around them. Another technology is SLAM, which allows the AMR to create a map of the environment and localize itself within that map in real-time. To ensure safety, AMRs are equipped with sensors that can detect humans and other obstacles in their path and can adjust their speed and direction to avoid collisions.

Additionally, open standards are being used to develop fleet management systems that can manage and coordinate mixed fleets of AMRs from different manufacturers. Customers are increasingly requiring that fleet management solutions be capable of running large and diverse fleets. These solutions can provide real-time data on the location, status, and performance of the AMRs, enabling operators to optimize fleet performance and efficiency. This enables interoperability between different types of AMRs, allowing them to work together seamlessly.

AI is expected to influence the next generation of AMRs by enabling them to make more intelligent decisions and adapt to changing environments. For example, AI algorithms can be used to optimize the AMR’s path and speed based on real-time data from the environment. AI can be used to predict and prevent potential collisions or accidents by analyzing data from the AMR’s sensors and other sources. Additionally, AI can be used to optimize material flows for manufacturing processes and managing inventory. Overall, AI will play a significant role in improving the efficiency, safety, and flexibility of AMRs in the future.

Joseph Alphonse: One of the latest emerging technologies for navigation for AMRs is visual SLAM navigation. This technology is a quantum leap over the current LiDAR-based sensors which are used to primarily detect and prevent collisions. Visual SLAM or VSLAM uses cameras to view and map the environment that the AMR is operating in. Through visualization, the AMR is able to accurately determine its location, view and detect obstacles, and make smart decisions such as modify its speed to optimize travel time as it reroutes around obstacles in a changing environment. Visual SLAM combines AI and 3D vision technologies to guarantee a superior performance in comparison to other guidance techniques for AMRs. Offering significant advantages over other forms of navigation such as magnetic tape, QR codes, and traditional 2D SLAM that require additional infrastructure to function, Visual SLAM AMRs are being embraced by companies to handle an expanding range of production and distribution tasks.

Tech Briefs: The common perception is that robots will take over jobs. What advice do you have for operations teams looking to implement AMRs, especially related to enabling workers’ acceptance of automation in their operations?

Image: ABB

Kevin Dumas: Robots do not take over jobs. In fact, our robots substitute or supplement the “dull, dirty and dangerous” tasks that people no longer want to perform. As for advice for operations teams looking to implement AMRs, we recommend that you: Communicate your automation plans in good time; Make your employees part of the process is the best way to smoothening your upcoming path to automation; Make the process enjoyable for your employees. AMRs can take on the most repetitive and heavy tasks, allowing employees to focus on high-value activities. Help your employees see that the robots are a tool for them to perform even better in their jobs. Let your employees know that the future of work is not robots that will replace them, but tools to help and work alongside people, increasing efficiency and safety. For example, AMRs take over manual tasks that are usually met with high absences due to work injuries. Show your employees that the robots will help them have better health and better job results.

Steve Branch: Long before AMRs arrive at the warehouse, Locus works closely with the on-site warehouse team on change management. For successful AMR integration, it’s important to communicate with employees about the role of automation as a resource that helps improve productivity while also making the job itself easier. Rather than replacing jobs, AMRs are a tool that enhances the worker’s job by doing the repetitive, physically demanding tasks. AMRs free up workers’ time to focus on more strategic, rewarding tasks. In addition, it increases worker retention and improves the overall workplace quality.

Jay Judkowitz: This is simply not the case today and won’t be for a long time if ever. Right now, there are 2 million manufacturing jobs in North America that are going unfilled. Robots can fill those jobs. And, where a driver is replaced, our customers are not laying that person off. They are training those people to do more skilled, more fulfilling, and higher paying work. Plus, all of the other workers who work near material handling but are not material handlers themselves find that they now have a safer, more injury-free environment which leads to greater comfort and job satisfaction.

Ed Volcic: It is worth differentiating between the effect of automation on jobs on a macro level vs a micro level. While the use of robots, including AMRs is on a fast rise year after year, the unemployment keeps on plummeting. On a micro level, on the other hand, a worker only considers the impact of automation on their own job. Hence the perception that robots will take over jobs is a concern among workers, and it is important for teams to address this issue when implementing AMRs. The following advice can help enable workers acceptance of automation: Involve workers in the implementation process; Communicate the benefits of AMRs; Highlight the role of workers in the new system; Provide opportunities for upskilling; Address concerns about job displacement. Overall, it is important for operations teams to involve workers in the implementation process, communicate the benefits of AMRs, and address concerns about job displacement to enable workers’ acceptance of automation in their operations.

Joshua Alphonse, ABB

Joseph Alphonse: No robot can replace a person. But often people are forced into dangerous, or repetitive tasks that robots could do. AMRs can handle ergonomically stressful tasks that free people from dull, dirty, dangerous, and repetitive tasks, enabling them to develop their skills and pursue more fulfilling jobs within a specific organization. It doesn’t take long for employees to see this, and the acceptance of the robotic support is an easy sell. With the tight labor market many companies are deploying robots to handle jobs they can’t fill with manual labor. They work hard to retain the valuable employees they do have, moving them to jobs with more upside potential. The implementation of AMRs provides a list of benefits from providing a safer work environment, more predictable lineside supply, and the realignment of resources to conduct tasks that increase the productivity of the manufacturing space or warehousing environment. The organizational benefits are widespread.

Tech Briefs: What are the three key technology trends that will shape the future of AMRs?

Kevin Dumas: As mentioned previously, three key technologies — 5G, advances in fleet management software and interoperability, and AI — will help shape the future of AMRs. As enterprises transition to 5G wireless networks, which are built specifically to move even more data more quickly and more accurately, AMR performance will also make a huge leap, with faster, more precise decisions about the paths they take within the facility. With facilities now deploying large numbers of AMRs, fleet management software with a single internal logistics system is even more important for coordinating, controlling, and managing the fleets. From a web-based interface, fleet management software improves the efficiency of AMR operations, enabling easy programming and control of the robots, including those with different top modules, hooks, or other accessories. As AI advances, it will continue to allow robots to be more effective at learning from what they experience to better detect and make intelligent choices in new situations. The autonomous behavior of AMRs will only continue to improve.

Image: Locus Robotics

Steve Branch: Three trends shaping the future of AMRs include increased AI capabilities for autonomous decision-making; Integration of IoT for real-time data exchange and coordination; and Advanced machine learning techniques for improved adaptability and learning from the environment. Locus’s commitment to operational improvement using AI has been a central part of the company’s DNA from the very beginning. As AI technology continues to improve, Locus will be leveraging these resources to continue to advance AMR technology for the warehouse. Locus is also committed to interoperability with other systems. To be effective, AMRs need to interface with other common types of warehouse automation such as conveyors, sortation systems, robotic arms, and more.

Jay Judkowitz: Three key trends we see are improvement in sensor technology: Right now, the only affordable safety rated technology is 2D planar LiDAR. Once we have cost-effective safety rated 3D sensors, safety, and localization will be able to be implemented at the highest levels at a reasonable price; digital twins and overall improvement of simulation technology: Right now, simulation tends to be an add on service from AMR vendors if it’s offered at all. But, as simulation technology improves in terms of quality and usability, that can be integrated as a standard part of fleet management and enable customers to continuously optimize their deployments and improve factory efficiency and ROI for the robots; adoption of cloud for data and analytics: Cloud-based offerings for storing and analyzing robot data can further help customers figure out how to optimize their material handling.

Ed Volcic: There are several key technology trends that will shape the future of AMRs by enabling them to become more intelligent, flexible, and efficient, and to operate safely alongside humans in a wider range of environments. Three key technology trends include: AI will play a significant role in the future of AMRs by enabling them to make more intelligent decisions and adapt to changing environments; advanced sensors such as LiDAR, 3D cameras, and ultrasonic sensors will continue to improve the navigation and perception capabilities of AMRs. These sensors will enable AMRs to detect objects and obstacles in real-time, navigate around them, and work safely alongside humans; cloud computing will enable AMRs to access real-time data and analytics from the cloud, allowing them to make more informed decisions and operate more efficiently; 5G network technology is expected to have a significant impact with AMRs of the future by supporting faster data transmission, lower latencies: or delays, in data transmission. 5G technology also provides improved reliability and network coverage, enabling AMRs to operate more reliably in challenging environments such as factories and warehouses.

Joseph Alphonse: Top technology trends which will continue to make autonomous robot implementations more efficient are the advancement of AI with respect to navigation and route management optimization, the decreasing cost of sensor and cameras, and the improvement of software tools for ease of commissioning and fleet management.

For more information, visit: