IFR: five trends will drive robot growth through 2025
Robot installations will grow quickly due to artificial intelligence, humanoid shapes, sustainable manufacturing, new business models, and labor shortages.
That is important because the increased use of robots has the potential to significantly reduce the impact of labor shortages in manufacturing, IFR said. That will happen when robots automate dirty, dull, dangerous or delicate tasks – such as visual quality inspection, hazardous painting, or heavy lifting—thus freeing up human workers to focus on more interesting and higher-value tasks.
To reach those goals, robots will grow through five trends in the new year, the report said:
1 – Artificial Intelligence. By leveraging diverse AI technologies, such as physical, analytical, and generative, robotics can perform a wide range of tasks more efficiently. Analytical AI enables robots to process and analyze the large amounts of data collected by their sensors. This helps to manage variability and unpredictability in the external environment, in “high mix/low-volume” production, and in public environments. Physical AI, which is created through the development of dedicated hardware and software that simulate real-world environments, allows robots to train themselves in virtual environments and operate by experience, rather than programming. And Generative AI projects aim to create a “ChatGPT moment” for Physical AI, allowing this AI-driven robotics simulation technology to advance in traditional industrial environments as well as in service robotics applications.
2 – Humanoids.
Robots in the shape of human bodies have received a lot of media attention, due to their vision where robots will become general-purpose tools that can load a dishwasher on their own and work on an assembly line elsewhere. Start-ups today are working on these humanoid general-purpose robots, with an eye toward new applications in logistics and warehousing. However, it remains to be seen whether humanoid robots can represent an economically viable and scalable business case for industrial applications, especially when compared to existing solutions. So for the time being, industrial manufacturers are still focused on humanoids performing single-purpose tasks only, with a focus on the automotive industry.
3 – Sustainability – Energy Efficiency.
Compliance with the UN's environmental sustainability goals and corresponding regulations around the world is becoming an important requirement for inclusion on supplier whitelists, and robots play a key role in helping manufacturers achieve these goals. In general, their ability to perform tasks with high precision reduces material waste and improves the output-input ratio of a manufacturing process. These automated systems ensure consistent quality, which is essential for products designed to have long lifespans and minimal maintenance. In the production of green energy technologies such as solar panels, batteries for electric cars or recycling equipment, robots are critical to cost-effective production. At the same time, robot technology is being improved to make the robots themselves more energy-efficient. For example, the lightweight construction of moving robot components reduces their energy consumption. Different levels of sleep mode put the hardware in an energy saving parking position. Advances in gripper technology use bionics to achieve high grip strength with almost no energy consumption.
4 – New Fields of Business.
The general manufacturing industry still has a lot of potential for robotic automation. But most manufacturing companies are small and medium-sized enterprises (SMEs), which means the adoption of industrial robots by SMEs is still hampered by high initial investment and total cost of ownership. To address that hurdle, Robot-as-a-Service (RaaS) business models allow enterprises to benefit from robotic automation with no fixed capital involved. Another option is using low-cost robotics to provide a “good enough” product for applications that have low requirements in terms of precision, payload, and service life. Powered by the those approaches, new customer segments beyond manufacturing include construction, laboratory automation, and warehousing.
5 – Addressing Labor Shortage.
The global manufacturing sector continues to suffer from labor shortages, according to the International Labour Organisation (ILO). One of the main drivers is demographic change, which is already burdening labor markets in leading economies such as the United States, Japan, China, the Republic of Korea, or Germany. Although the impact varies from country to country, the cumulative effect on the supply chain is a concern almost everywhere.
Picture a busy DC, with manually operated forklifts, people, and pallets in constant motion. At the same time, the stationary equipment they interact with, such as conveyors and palletizers, is industriously whirring away. Together, they are performing something akin to a carefully choreographed ballet.
Now add driverless forklifts to the mix. Shuttling along without a human operator on board, they may look like they’re operating independently, but they’re not. They’re actually in constant contact with other equipment and software, making sure they perform their part in the dance at the right moment. Without that ability to communicate, the forklifts—and other warehouse operations—could come to a standstill.
Who, exactly, are driverless forklifts “talking” to, what information are they sharing, and how does that exchange happen? We asked automation experts to explain. They also shared tips on ensuring successful communication between automated lift trucks and other equipment and software.
TWO-WAY COMMUNICATION
Lift trucks that do their jobs without a human operator on board cannot “speak” directly to each other. “As it stands now, there is no peer-to-peer communication or interaction on a forklift-to-forklift basis,” notes David Griffin, chief sales officer for Seegrid, a developer of autonomous lift trucks and AMRs (autonomous mobile robots). There is, however, interaction between forklifts via a centralized fleet manager system (also referred to as a traffic management system or an automation server). This “overarching conductor of the automated system” assigns tasks to each forklift, controls the route the trucks will follow, and manages traffic flow, says Nick McClurg, a sales engineer at forklift maker Hyster Co.
The forklifts communicate with many kinds of material handling equipment, such as robotic palletizers and depalletizers, stretch wrappers, conveyors, automated storage and retrieval systems (AS/RS), and dock equipment. That communication must be bidirectional, says Michael Marcum, senior director of autonomous vehicles at systems integrator Bastian Solutions, a Toyota Automated Solutions company that also makes robotic forklifts. Much of the exchange consists of messages that indicate status—whether or not the two pieces of equipment involved are ready to conduct a transaction. For example, if a forklift will be delivering a pallet to a stretch wrapper, then the wrapper has to tell that forklift, via the fleet manager system, that the load position is empty and the forklift is allowed to set a payload there, Marcum explains. After a pallet has been wrapped, the stretch wrapper will call for a pickup via the fleet manager. Once the forklift picks up the wrapped pallet, it must confirm to the stretch wrapper that it has departed; without that signal, the wrapper cannot receive its next load.
If a truck is not ready for an assigned task, it signals that status to the fleet manager, and the task will be reassigned to another nearby vehicle, says Jayce Nelson, sales manager, North America, for Kion Group’s Linde Automated Solutions, a specialist in automated forklifts and software. When the assigned forklift is ready to approach, say, the end of a conveyor to pick up a load, it uses its vision systems, such as 3D cameras, to align itself with the equipment.
With their control software, robotic forklifts also have the ability to communicate with other warehouse equipment, like fire-detection systems and automated rollup doors. “If a device is capable of sending or receiving electrical signals, then the vehicle can interact with that device via the automation host software,” McClurg says. Even a piece of mechanical equipment could be outfitted with sensors that help it interact with automated forklifts, according to Brian Markison, director of sales for Rocrich AGV Solutions, a joint venture of Mitsubishi Logisnext’s Rocla and Jungheinrich units that specializes in automated guided vehicles.
The capability to communicate with different types of devices enhances warehouse safety, Griffin says, because it enables automated forklifts to talk to safety equipment like pedestrian warning lights and intersection gates. And since the robots constantly transmit their location, the traffic control system can identify developing problems and prevent them. For example, the system will stop an autonomous forklift from entering an intersection that’s occupied by another lift truck. Once the other truck has moved on, the system will give the approaching forklift the “all clear,” he says.
Hardware isn’t the only thing driverless forklifts can talk to; they also are in continual dialogue with various types of software. “Most commercial warehouse software programs today have the capability for two-way communication, and most can be integrated with automated lift truck fleet management software,” observes John Wilkins, a sales engineer for Yale Lift Truck Technologies. The most common are warehouse management systems (WMS) and warehouse control systems (WCS); others include enterprise resource planning (ERP) software, fleet management and telematics systems, and transportation management (TMS) and order management (OMS) systems.
As for how that might work, Rocrich’s Markison gives the example of a WMS sending an order to move a pallet from one location to another. The order typically will include start and completion time, and some indication of the move’s priority. “That order can then be taken into the fleet manager, which will appropriately queue up the tasks that need to be done,” he explains. The forklift must also report completed missions back to the WMS.
HOW TO TALK TO A FORKLIFT
Communication between robotic forklifts and warehouse equipment and software happens in a number of ways. Which method is deployed depends on the equipment and software involved as well as the tasks to be carried out. Each installation is unique in some way, but there are some commonly used approaches.
Some communication protocols are more widely used than others. Examples of those in widespread use include modbus, a serial communication protocol that governs an initiating and a responding device, and CANbus (Controller Area Network), a real-time communication protocol that transmits data to networked industrial controls.
A driverless forklift’s interface with other equipment could be something electromechanical, such as a photo-eye sensor, says Jeff Kuss, product manager–automated solutions at forklift maker and intralogistics specialist The Raymond Corp. A sensor at the end of a conveyor, for instance, could detect the presence of a pallet. That triggers the sensor to create an electrical signal that it sends to a programmable logic controller (PLC). The PLC receives the electrical signal as a digital input and then transmits a message, via ethernet, to the server that controls the automated vehicles. Finally, the server sends the instructions over Wi-Fi to the closest available forklift to “pick up the pallet and take it to Location X.” (Some facilities use Bluetooth or cellular transmission instead of Wi-Fi.) Data that identify loads and trigger a task can also be acquired through IoT (internet of things) platforms, RFID (radio-frequency identification) systems, and barcode scanning.
Another option, Bastian’s Marcum says, is to use infrared-based optic couplers that share bits (binary digits, the smallest units of digital information) as inputs and outputs. When the forklift gets within a certain distance of another piece of equipment, “the two devices can talk to each other, similar to the way a TV remote works,” he explains.
Usually, though, software is a critical intermediary between driverless forklifts and other equipment. It can be complicated. In the case of a WMS, McClurg says, his company’s approach is to send a text or JavaScript Object Notation (JSON) file to the WMS; in exchange, the WMS sends a file to a folder on the localized network that can be accessed by the automation host software. The fleet manager reads the file and executes it. Once the task has been completed, a message is sent to another folder. The WMS opens it, reads it, and, based on its contents, either closes out the order or sends additional instructions.
To ensure that interfacing software programs understand the messages they receive from each other, it’s often necessary to create an application programming interface (API). An API is a comparatively simple type of middleware—or software layer—that acts as a translator, facilitating communication by reformatting messages so they will be intelligible to the receiver. In essence, they are “setting ground rules in terms of what information is passing back and forth and what it means,” Markison explains.
In some cases, more complex middleware may be needed. According to Brice Bucher, senior manager of products at software developer and systems integrator Flexware Innovation, APIs have limitations. In a presentation at the Autonomous Mobile Robotics & Logistics Conference 2024, Bucher noted that APIs don’t address data transformation, protocol conversion, or business logic integration. When each system has different data formats or requires specific protocols, middleware bridges those gaps, he said. Middleware also ensures that data moves between systems without delay, he said. For example, if an AGV completes a task, middleware can instantly trigger updates across systems, so that WMS, ERP, and other systems are aligned in real time.
CAN WE TALK?
Raymond’s Kuss notes that each communication integration will be unique in some way. That’s partly because automated forklift vendors and suppliers of fleet manager systems have proprietary interfaces. On top of that, software with some degree of customization, such as a WMS, may require modifications to the fleet manager system, he explains. What’s more, adds his colleague John Rosenberger, director, iWarehouse Gateway & Global Telematics, “even if we know the format for efficient data transfer, the content of the messages may differ depending on the forklift manufacturer, or it can be different by functionality.”
Mixed fleets with forklifts from different manufacturers present a particular challenge. Seegrid’s Griffin notes that it’s common for facilities to use robots from multiple vendors. Generally, he says, each automated solution has its own proprietary fleet manager software that understands where all units under its purview are and controls their movements. When robots of different brands cannot be confined to separate areas, it’s important that their fleet managers have the ability to communicate, so they can do things like open and close intersections where different types of robots cross paths.
While it is possible for dissimilar fleet managers to talk to each other, that’s easier said than done. “Those systems inherently are not interoperable,” Nelson says. “The need to share information like coordinates, current status, past assignments, and prioritization makes it difficult to assign travel paths.” In addition, if the forklifts are unable to communicate location information and what they are doing, that can lead to deadlock, where the vehicles simply stop—what Yale’s Wilkinson calls “the classic situation: a staring contest between two autonomous vehicles from different OEMs, neither one capable of blinking or losing.”
A solution for some facilities is third-party fleet manager software that’s designed to work in multiple brands of autonomous forklifts; examples include those offered by independent developers such as Kollmorgen, BlueBotics, Navitech Systems, and Flexware Innovation. In fact, some forklift OEMs partner with these and other providers instead of developing their own fleet managers. This opens the way for a fleet to potentially buy different robots utilizing the same control and navigation system, which will reduce complexity to some degree, Marcum says.
Communications with driverless forklifts may become simpler in the near future. VDA 5050, an open-source protocol for communication between AGVs and fleet manager systems, is currently in development. Coordinated by two German industry organizations, one for auto manufacturers and the other for material handling and intralogistics, this universal protocol promises to allow “any mobile robot, regardless of brand, [to] be seamlessly integrated into existing operations,” wrote Alfredo Pastor Tella, who runs the Europe-based AGV Network website, in a LinkedIn post. Pastor Tella wrote that Kollmorgen will introduce VDA 5050 into its robot control software in 2025, but other industry observers have noted that because the protocol’s roots are in European manufacturing and there are still technical issues to be worked out, it may be a few years before it takes hold in the forklift world. When it does, conversations with autonomous forklift fleets will likely become much less complicated to hold.
Tips for success
Want to be sure your driverless forklifts will always “get the message”? Here are some experts’ recommendations for facilitating communication with them:
Involve your IT experts early! They’ll need to identify what relevant data is currently available and where it resides. Make sure they’re comfortable that any APIs and other software meet your company’s security requirements. For cloud-based systems, verify that the vendor and systems integrator will have remote access if they need to service any of the systems or software. (Brian Markison, Rocrich AGV Solutions)
If you’re buying from different manufacturers, find out which supplier has navigation technology on the brands you’re considering and try to stay with a single system if possible. If you have a single platform, you can make a change just once and the entire fleet will receive that modification. If you have two fleet managers, segregate them as much as possible. Wherever they are separate, you’ll only have to change that one, but in shared areas, you’ll have to change both. (Michael Marcum, Bastian Solutions)
When it comes to facilitating communication, software is not always the best answer. Sometimes something simpler, like PLCs that notify equipment through very basic logic, works just fine. And it’s better to start small and integrate each function as you go, rather than try to integrate everything at once. You can tie two systems together and demonstrate the benefits from that, then use the savings to justify and help fund the next piece. (John Rosenberger, The Raymond Corp.)
Conduct testing in real-world scenarios, and make sure legacy software and communication technologies are compatible with the automation. These systems work in a dynamic environment, and a lot changes over time. Calibration tests can make sure everything still aligns correctly. And remember to inform your vendor of changes in things like throughput rates, layout, pallet sizes and configurations, products, and so on. (Jayce Nelson, Linde Automated Solutions)
In most facilities, commands and data are communicated via Wi-Fi, so connectivity and reliability are a top concern. A pre-installation survey to measure Wi-Fi signal availability and strength throughout the facility is an absolute must. Based on those findings, you may need to enhance signal strength and expand capacity and coverage. In some very large facilities, a private wireless network that uses cellular signals may be the best solution. (Deryk Powell, CEO, Velociti Inc., a provider of technology deployment, support, and integration services)
The deal will add the Google DeepMind robotics team’s AI expertise to Austin, Texas-based Apptronik’s robotics platform, allowing the units to handle a wider range of tasks in real-world settings like factories and warehouses.
The Texas firm joins other providers of two-legged robots such as the Oregon company Agility Robotics, which is currently testing its humanoid units with the large German automotive and industrial parts supplier Schaeffler AG, as well as with GXO. GXO is also running trials of a third type of humanoid bot made by New York-based Reflex Robotics. And another provider of humanoid robots, the Canadian firm Sanctuary AI, this year landed funding from the consulting firm Accenture.
“We’re building a future where humanoid robots address urgent global challenges,” Jeff Cardenas, CEO and co-founder of Apptronik, said in a release. “By combining Apptronik’s cutting-edge robotics platform with the Google DeepMind robotics team’s unparalleled AI expertise, we’re creating intelligent, versatile and safe robots that will transform industries and improve lives. United by a shared commitment to excellence, our two companies are poised to redefine the future of humanoid robotics.”
David Scheffrahn is the North American vice president of sales at Ocado Intelligent Automation, a part of the technology specialist Ocado Group. Although he began his career focusing on robotic solutions for semiconductor, electronics, and automotive manufacturers, Scheffrahn eventually moved on to the logistics sector, where he worked at Rethink Robotics, Seegrid, Plus One Robotics, and Dexterity before joining Ocado in 2023. He holds a degree in mechanical engineering from the University of Texas.
Q: How would you describe the current state of the automation industry?
A: Today, automation is available for nearly every task in the supply chain. Yet we know from industry analysts that only one-fourth of warehouses are “automated.” [The market research firm] Interact Analysis predicts that 27% of warehouses will be automated by 2027.So many warehouse operators still have the opportunity to embrace and benefit from automation.
Whether companies are just getting started with automation and could benefit from swapping out manual carts for automated ones or are looking for an end-to-end omnichannel fulfillment solution, there will be options available.
Q: You’ve worked in the robotics industry for the past 25 years. What changes have you seen in robotic design and applications during that time?
A: Believe it or not, robots pre-date me! I fell in love with robots right out of college. When I graduated in 1994, I was hired by a local robotics company, and one of my early jobs was to program robots to cut circuit boards into the correct shape to fit into cellphone housings. I was hooked for life. Back then, robots did exactly what you programmed them to do, very precisely, over and over.
In the mid-2000s, an explosion of software and sensor-based technologies started to give robots the capability to operate in environments that are much less structured, such as warehouses and fulfillment centers. Nowadays, robots can perform a wide range of tasks and movements, seemingly on the fly. They can interact with the world around them—and even people—because they can safely operate and adapt to changes in the environment.
Q: How are artificial intelligence and machine learning being applied to robotics?
A: Think of a robotic pick arm. Traditionally, it was trained and tested to always pick the same—or very similar—object or item set. Now, when we apply artificial intelligence, vision systems, and sensors to the same robotic arm, it can teach itself to handle new items without previous training or testing. Vision systems and sensors scan shapes and identify items to direct the arm on how to handle fragile products without damaging them or how to grasp an item with a new and different shape.
Q: Automation used to be a major investment. Has it become any easier for smaller companies to get started with automation?
A: A few years ago, automating was a choice. In 2024, the question isn’t whether you should automate, but rather what’s the right automation solution for your operations. Automated solutions can be big or they can be small, but they should always improve warehouse operations and be “right-sized” for the application.
Autonomous mobile robots (AMRs) are some of the most approachable automated solutions available for 3PLs or small and mid-sized warehouses. AMRs can be deployed quickly one at a time or by the dozen. They can integrate seamlessly with existing warehouse systems and infrastructure, and work safely alongside human pickers. Customers we have worked with report that deploying automated carts based on AMRs has doubled their productivity, improved accuracy by 40%, and reduced employee training time by 80%.
Q: What is the next frontier in robotic design and applications?
A: The use of 3D printing is opening up new opportunities in robotic design. I think we’ll see that technique used more because of the resulting benefits.
Robots made via 3D printing are lighter, which, in turn, means the grids used in automated storage and retrieval systems (AS/RS)—like the Ocado Storage & Retrieval System (OSRS)—can be lighter. Lighter grids are easier and quicker to assemble. But more importantly, in Ocado Intelligent Automation’s solution, they can provide 33% more vertical storage capacity within the OSRS than heavier grids. The more cubic density in an AS/RS, the more warehouse operators can conserve footprint, lower real-estate costs, and scale inventory.
Q: How is Ocado Intelligent Automation expanding its offerings for the supply chain industry?
A: Ocado Group has been developing automated technology for more than 20 years. In 2023, it formed Ocado Intelligent Automation (OIA), the division I work in, to bring automation solutions to intralogistics (supply chain activities that take place within a warehouse) and to sectors beyond online grocery, which is where the company got its start.
Online grocery is one of the most demanding e-commerce environments—with needs that are very analogous to the fulfillment and logistics requirements of the health-care, retail, consumer packaged goods, and third-party logistics sectors. I can’t wait to see how these sectors benefit from OIA technology and robotics in the coming years. It’s going to be impressive!
Airbus Ventures, the venture capital arm of French aircraft manufacturer Airbus, on Thursday invested $10.5 million in the Singapore startup Eureka Robotics, which delivers robotic software and systems to automate tasks in precision manufacturing and logistics.
Eureka said it would use the “series A” round to accelerate the development and deployment of its main products, Eureka Controller and Eureka 3D Camera, which enable system integrators and manufacturers to deploy High Accuracy-High Agility (HA-HA) applications in factories and warehouses. Common uses include AI-based inspection, precision handling, 3D picking, assembly, and dispensing.
In addition, Eureka said it planned to scale up the company’s operations in the existing markets of Singapore and Japan, with a plan to launch more widely across Japan, as well as to enter the US market, where the company has already acquired initial customers.
“Eureka Robotics was founded in 2018 with the mission of helping factories worldwide automate dull, dirty, and dangerous work, so that human workers can focus on their creative endeavors,” company CEO and Co-founder Pham Quang Cuong said in a release. “We are proud to reach the next stage of our development, with the support of our investors and the cooperation of our esteemed customers and partners.”
Tire manufacturer Michelin has long used predictive maintenance tools to head off equipment failures, but the company recently upped its game by implementing cutting-edge robotics at its factory in Lexington, South Carolina. Managers there are using Boston Dynamics’ autonomous mobile robot (AMR) “Spot” to speed and streamline the inspection and maintenance processes—a move that is boosting productivity at the Lexington facility and for the company at large.
“Getting ahead of equipment failures is important, because it affects our production output,” Ryan Burns, an associate in the facility’s reliability and methods department, said in a case study describing the project. “If we can predict a failure and we can plan and schedule the work to fix the issue before it becomes an unplanned breakdown, then we’re able to increase our output as a company and a tire producer.”
MORE—AND BETTER—INSPECTIONS
Spot is a versatile quadruped AMR that can automate sensing and inspection tasks, and capture data—all while moving freely throughout a facility. The robot is being used around the world for maintenance-related functions, such as detecting mechanical problems and monitoring equipment for energy efficiency. At the Michelin plant, managers began by assigning Spot to inspect machinery in its tire verification (TV) area—taking over tasks previously done by in-house technicians as well as conducting additional inspections. Spot identifies issues and problems, and then conveys that information through its software program, called Orbit, which managers can access via an on-site server. From there, managers can sort through the data to detect anomalies and set alarm thresholds that will trigger a technician’s response.
“From a technician standpoint, Spot going out and doing these routes eliminates a mundane task that the humans were doing,” said Burns. “By Spot finding these anomalies and these issues, it gives the technicians more time to … [decide] how and when they’re going to fix the problem versus going out, identifying [the issue], then trying to plan and schedule everything.”
FEWER BREAKDOWNS, MORE PRODUCTIVITY
The results have been game-changing, according to Burns and his colleague Wayne Pender, the tech methods and reliability manager at the Lexington plant. As of this past fall, Spot was running seven inspection missions in the TV area, scanning about 350 points across 700 assets to detect anomalies ahead of time. The results helped generate 72 work orders in Michelin’s system—allowing the facility to avoid uncontrolled breakdowns and major production losses, according to Pender. On top of that, Spot had generated 66 air-leak work orders, identifying areas where Michelin can reduce energy consumption.
Looking ahead, the plan is to apply Spot’s talents beyond the TV area to the rest of the facility.
“Spot is a member of our maintenance team,” Burns said. “The future is to have more Spots, so that we can improve on our inspections and improve our overall output as a company here at [Lexington].”
Pender agrees: “We see Spot [as] the future. … [But] we probably need a whole dog pound or multiple Spots … to actually do what we need to do [across all of Michelin’s North American facilities].”