Automated lift trucks—a.k.a. robotic, autonomous, or driverless forklifts—and their close cousins, automated guided vehicles (AGVs) and autonomous mobile robots (AMRs), have become a hot commodity. At trade shows like ProMat and Modex, more and more exhibitors are showcasing robotic trucks along with the traditional driver-operated vehicles. Joining them on the show floor are vendors of related technologies, such as Phantom Auto, the MHI Innovation Award winner that enables remote monitoring and operation of forklifts; and Third Wave Automation, which provides autonomous forklifts in collaboration with Clark Material Handling Co. as well as remote monitoring and operation.
Interest in this technology is largely being driven by two things: the ongoing difficulty of hiring and retaining skilled workers in DCs, and the broader move toward automation by companies seeking to optimize their operations at levels manual equipment can’t achieve.
If you’re among the growing number of decision-makers planning to introduce driverless forklifts in your DC, then some changes to infrastructure, operations, and staffing will likely be needed before you send them out onto the floor. Here are some things to consider as you get your DC “robot ready.”
Layout: An accurate picture of the facility’s layout will reveal areas that could spell trouble for automated forklifts. Start with an accurate, to-scale technical drawing showing all the details, recommends Martin Buena-Franco, manager, automation products marketing for The Raymond Corp. It should include the exact dimensions and locations of such features as aisles, walkways, and storage racks; infrastructure like pillars, ramps, and doors; lighting, electrical service, and charging stations; overhead obstructions; and functional areas, such as loading/receiving docks and packing stations, and related stationary equipment.
A common problem for robotic forklifts is insufficient aisle width. Automated vehicles are surrounded by a sensor-generated “safety field” (also referred to as a “sensor field”) that detects objects and pedestrians. If the aisle is too narrow and the racking falls within the safety field as the truck pivots or positions itself for a pickup or dropoff, the robot will interpret the racking as an obstacle and will stop. For that reason, automated trucks require more aisle width than a manual version of the same truck does, says Kyle Smart, sales manager, robotics, for Yale Lift Truck Technologies. The actual measurements depend on the type of forklift and its environment, but a rough estimate is an extra six to eight inches.
Too-narrow aisles can slow things down in other ways. As the vehicle moves along, the sensors’ field of view adjusts to the truck’s surroundings, says Jared Green, director of global sales–automation and emerging technology at Crown Equipment Corp. “When you have more space, the sensor field can be longer, and the truck can maintain speed,” he explains. “If you don’t have sufficient space, then the extent of the sensor field will be drawn in and the speed will automatically be adjusted downward.”
Facility layout can also create challenges for vehicle navigation systems. Consider LIDAR (light detection and ranging) technology, which provides object perception, identification, and avoidance, for example. A large open area with storage configurations that change frequently—such as floor-stacked loads that are moved into and out of the area—can be difficult for LIDAR to navigate, says Brian Markison, director of integrator and direct sales for Rocrich AGV Solutions, a collaboration of Jungheinrich and Mitsubishi Logisnext Americas’ Rocla AGVs. Another example: Laser-guided navigation utilizes light bouncing back from wall-mounted reflectors to precisely calculate location. Narrow aisles can prove challenging if a navigation “window” through the storage rack is not available and the reflection is partially or fully blocked. Decisions about which type(s) of navigation would be best suited for a particular facility should take such limitations into account. Markison recommends a site visit by the solution provider as the best way to identify potential navigation problems associated with your facility’s unique layout.
Condition: The facility’s condition—especially the state of the floors—matters because robots need flat, smooth surfaces to operate safely and efficiently. Automated vehicles can’t spontaneously adjust to variable floor conditions the way a human operator would, says Mick McCormick, director of intralogistics systems, North America, for Linde Material Handling, a unit of Kion North America. And while it’s possible to program slight shifts in travel paths to avoid rough patches, you won’t have such flexibility at pickup and drop locations, he adds. In some cases, it may be necessary to grind cement flooring to make it level and smooth before bringing in robotic trucks.
Material and information flow: To get the greatest benefit from introducing robotic forklifts to your DC, it’s necessary to consider the operation as a whole. “People interacting with automation changes the process of how product comes in, moves through, and goes out of a DC,” Crown’s Green observes. “It forces you to look at the entire operation and [determine] whether with automation, there are new and better ways to optimize things that maybe you couldn’t control as well with manual methods.”
That’s why a critical early step in preparing for automation is understanding the flows of equipment, material, and information. That includes all processes in the DC and how they interact with each other, Markison says. He recommends identifying each decision point and documenting how and when operators and equipment receive information, what roles they perform, and where they travel. With that baseline information in hand, you can work with the equipment supplier and systems integrator to map out any changes that will be required.
You may, for example, have to revise travel paths. Operating automated and manual forklifts in close proximity creates opportunities for an operator to inadvertently cross a robot’s safety field, causing it to stop, says McCormick. To forestall that and other human-machine conflicts (almost always caused by human error, in his experience), he recommends avoiding path overlap by isolating manual and robotic forklifts’ operating zones to the extent possible.
Physical handling protocols may also require adjustment. For example, process customization will negatively affect robots’ productivity, Smart says, so to maximize efficiency, you may want to store loads that require different handling processes in separate locations. In addition, pallet-quality issues, such as overhang, load lean, and skewed positioning in the racks, can make them hard for automated forklifts to handle. Establishing operating standards that eliminate such errors can be helpful in reducing the need for human intervention, he suggests.
But it’s not just physical handling that merits forethought; information flows may change too. For example, how a WMS (warehouse management system) creates tasks and distributes instructions will change when driverless forklifts are put into service, Green says. For that reason, you may want to allocate additional resources to planning and preprogramming, he says.
Technology considerations: In most facilities, commands and data are communicated to, from, and among automated forklifts via Wi-Fi, so Wi-Fi connectivity and reliability are a top concern. If the signal is unavailable at any decision point where the robot requires permission to go ahead, stop, turn, or perform some task, it will come to a stop, and any robots behind it will be affected, McCormick says. If the signal is weak, the trucks will slow down as they search for a stronger signal. A detailed pre-installation survey to measure Wi-Fi signal availability and strength throughout the facility is therefore an absolute must.
When planning upgrades or alternatives to your current wireless network, three areas bear close attention, advises Deryk Powell, president and COO of Velociti, a consulting and systems integration firm that manages implementations of automated forklift technology, among other services. They are as follows:
1. The technology on the vehicle itself. Truck-mounted or embedded technology can include tablets, handheld scanners, and other devices that connect to Wi-Fi; telemetry software; navigation systems, such as LIDAR, radar, and simultaneous localization and mapping (SLAM); collision-avoidance systems; RFID (radio-frequency identification) readers; and video cameras, to name a few examples. It’s important to understand their connectivity requirements, individually and in total.
2. Competition for capacity. In any given facility, the Wi-Fi capacity is limited. In consequence, prioritizing the automation platform can create conflicts with other users of the network, while overburdening a Wi-Fi system will degrade service for all users, including the robots. In such cases, separate Wi-Fi networks may be necessary. To make sure you have sufficient capacity, Powell advises, identify all IT initiatives planned for the next 24 to 36 months that could affect demand for Wi-Fi, and invest in a system that will be able to meet that demand.
3. Wi-Fi performance. Except for installations with a limited number of machines in a small area, the existing Wi-Fi network will likely be insufficient for the job at hand, Powell says. In many cases, signal strength can be enhanced, and capacity and coverage expanded. But in very large DCs with a lot of automated equipment, such enhancements may not achieve the required signal quality, availability, and reliability.
For those facilities, Powell says, an appropriate alternative may be a private LTE (long-term evolution) wireless network, which uses cellular signals and has its own radio-frequency spectrum. The advent of 4G LTE technology upped the performance of cellular networks to the point where they could compete with Wi-Fi in DC applications, he notes. The next generation, the 5G communications standard, is faster and offers capacity improvements over 4G, Powell says. But regardless of which standard is selected, devices using the wireless network must be equipped with the appropriate chips.
Private wireless networks are more expensive than Wi-Fi in terms of their initial cost, but in the right circumstances, the advantages they offer make them a worthwhile investment. According to Powell, their capacity is greater and they can support many more connections than “normal” Wi-Fi can. When properly designed, they have limited to no dead or weak spots, and offer superior performance with respect to the speed of communications and latency. (Latency is the lag between the time a signal is sent and the moment the machine takes action.) With Wi-Fi, he says, latency is from milliseconds up to a second or two, while in private wireless networks, the response time is measured in single-digit milliseconds.
Wireless communication is not the only IT concern when introducing automated forklifts. See the accompanying sidebar for other important considerations.
Charging and electrical: Driverless forklifts are directed to charging stations by fleet management software, typically when they’re not scheduled for any immediate tasks. Generally, they charge their batteries without human intervention, either by making physical contact with a charging station or via wireless charging.
Chargers should be located where the automated forklifts do their work. One reason, says John Rosenberger, director, iWarehouse Gateway and global telematics at The Raymond Corp., is that robotic forklifts typically move more slowly than trucks operated by drivers and move at a consistent pace. As a result, machines traveling across a facility to reach charging stations could be disruptive, especially if they have to cross zones where manually operated forklifts are in use. Installing multiple charging stations where the robots operate will reduce or prevent such disruptions. However, this may require running high-capacity electrical service to locations where you didn’t have it before.
Equipment maintenance: Robotic forklifts are still so new that it’s hard to quantify their impact on equipment maintenance. Yale’s Smart cautions that while it’s true that without human drivers, impacts will be rare to nonexistent and that there tends to be less wear and tear on things like tires, brakes, and drive-train components, there is a tradeoff: Robots generally run more hours than manual trucks do, often operating at 80% to 85% utilization, so the number and frequency of planned maintenance sessions will increase.
Furthermore, technicians who maintain automated forklifts need special training and skills, which raises the cost of maintenance, Rocrich’s Markison says. They must know how to clean and recalibrate sensors so they operate correctly, and they may also have to repair or replace sensors and cameras, and troubleshoot traffic management software.
Maintaining automated forklifts can be so complex that “layers of support” are called for, Crown’s Green says. His company, for instance, not only deploys automation-trained technicians at the local level but also provides centralized support by automation and software engineers as well as additional field engineering resources.
Safety protocols: Before robots get to work, all personnel—including those who don’t directly interface with them—must receive training on safety protocols that are specifically designed for working around AGVs, Markison says. For example, although the forklifts usually travel slowly, they may be programmed to go faster on a straightaway; at full speed, “it takes them a certain amount of distance to slow and stop, so everyone must have an understanding of where a person should and should not be” relative to the robots’ travel path, he explains.
Many companies require pedestrians to stop and make eye contact with the forklift operator, then wait for an acknowledgment before going ahead. But with robots, “you don’t have that, so you need protocols for various ways of interfacing with them,” Smart points out. For instance, when an automated forklift has to stop unexpectedly, such as for an in-transit adjustment, workers could place an obstacle in the aisle to prevent other robotic traffic from entering until the problem has been resolved.
GETTING EVERYONE ON BOARD
When automation is on the horizon, employees (understandably) worry about their jobs. They need to know whether they will lose their jobs, and if they keep them, how their jobs will change, what new career opportunities will open up, and what kind of personal financial impacts there could be.
Ideally, forklift operators wouldn’t lose their jobs; instead, their work would be redefined. For instance, they could focus on areas that require judgment and decisions, letting the machines handle routine tasks like horizontal travel. They could also fill open positions in other areas of the DC or train for more complex, value-added jobs, Raymond’s Buena-Franco suggests. An example of the latter is “robot wrangler” or automation specialist—a key, new position on each shift that involves managing the automated fleet and resolving problems.
Frequent, ongoing communication—about not just what will change, but also why—will help get people at all levels on board. “You have to justify the project, of course, but then you have to flow that justification down into the organization,” Raymond’s Rosenberger says.
And it’s not enough to simply talk about it. “If you have shown how it is going to work, using existing warehouse operations as examples, and actually acted on it to prove that the benefits can be activated,” he adds, “that will build trust within the organization, which is crucial for the acceptance of automation.”
Surprise! Tech issues you may not have thought about
- While the strength and reliability of Wi-Fi or private communication networks usually are top of mind when planning for robotic forklift implementations, other critical tech issues also require close attention. Here are three you might not have thought about:
- Managing data flows. It may be necessary to modify your WMS and/or ERP (enterprise resource planning) system so it can communicate with the forklift control system via an automated web service interface (AWI), says Jared Green, director of global sales–automation and emerging technology at Crown Equipment Corp. If your IT group does not have that capability, then you will likely need a custom middleware integration by the forklift vendor, its technology partner (Crown, for instance, collaborates with AGV maker JBT), or a third-party systems integrator.
- IT updates. Companywide IT decisions sometimes have unintended consequences for local equipment, says Mick McCormick, director of intralogistics systems, North America, for Linde Material Handling. A real-life example: Automated forklifts in a DC mysteriously stopped working. An investigation revealed that a security-related update mandated by corporate IT had inadvertently caused the Wi-Fi to stop recognizing the robots’ IP addresses. As this episode suggests, it’s a good idea to establish a Plan B for communicating with automated forklifts.
- Security. Every connection to a network is a potential way in for cybercriminals, according to John Rosenberger, director, iWarehouse Gateway and global telematics at The Raymond Corp. For security’s sake, every connected device on lift trucks, remote server connection, application programming interface (API) portal, and other connection should come with proof that it was tested by an independent third party to verify compliance with approved security standards—before it can be integrated with any system that connects to the network, he says.
