Autonomous mobile robots (AMRs) and larger automated guided vehicles (AGVs) have the potential to significantly increases an organization's efficiency.
This potential can be maximized if the right robotic solution is chosen, and if that vehicle is then commissioned to perform a task that it suits. Conversely, choosing the wrong system can lead to frustration and delays, and it could impact the potential return on investment that can be achieved.
AMR and AGV technology can deliver great productivity and efficiency benefits, as well as improving the safety of your site. To optimize these benefits from the very first rollout and increase your chances of investment success, be sure to avoid the following four mistakes.
Also read: Industrial robotics receive room to grow at Fraunhofer IPA
Figure 2: An AMR or AGV can only do as it is programmed, so your business' processes should be analyzed, simplified and standardized to suit your robot's specific capabilities.
Mistake 1: Trying to automate a process as is
Successful AMR and AGV projects do not simply take existing manual processes and automate them (Figure 1). Automation is a journey. Therefore, when adopting mobile robots, take the opportunity to re-evaluate your processes.
An AMR or AGV can only do as it is programmed, so your business' processes should be analyzed, simplified and standardized to suit your robot's specific capabilities (Figure 2).
The level of changes to a process will depend upon the application. For example, warehouses are usually designed with transportation in mind, so, in general, they will lend themselves to automation with only minor changes.
In manufacturing plants, on the other hand, material intralogistics is usually a support function, and the layout and infrastructure of the site might require bigger changes to accommodate AMRs or AGVs.
One example of best practice comes from GECOM, a U.S.-based automotive parts manufacturer headquartered in Greensburg, Indiana, that decided to optimize its material-handling resources by deploying five automated Nipper pallet trucks by F3-Design.
Recognizing that the site's existing intralogistics processes and environment—narrow aisles—had not necessarily been designed for automation, its team started preparing its AGV program by deploying small, manually driven vehicles in place of human material-handling staff. This approach enabled GECOM's team to establish vehicle routes and pick-and-drop points.
Team leaders would generate missions using the same tablets and site-management software used by the material-handling associates who were driving the vehicles. This gave the AGV team a clear picture of the processes involved, which the AGVs would need to take over, enabling a smooth transition when the AGVs arrived, as these were simply slotted in to replace the human-driven vehicles.
Mistake 2: Not checking accuracy and robustness
AMRs and AGVs must not only be reliable and cost-effective, but, for the majority of applications, they also require a good level of accuracy and flexibility.
For example, an AMR whose job is to move beneath racks to pick up and drop off light payloads might need a positioning accuracy of approximately 1 cm, or a half inch. It will also need to be able to dock accurately with battery-charging stations in order to ensure hands-free, continual operation.
If a system has inferior positional accuracy, which is largely determined by the type and quality of its built-in navigation system, it might need an additional line-following system or fixed conveyor adding at pick-and-drop points to ensure the robustness of its operation. This dual approach will result in slower speeds, slowly and more costly vehicle installation projects and lower levels of efficiency overall.
Investing in a robotic vehicle that is relatively untested could lead to its not functioning as planned, leading to reduced availability and increased maintenance overheads, both of which will impact the system's potential ROI.
To reduce risk, it is prudent to investigate how many of a manufacturer's mobile robots have been successfully deployed and are in current operation. If you can, take the time to speak to existing users of your preferred solution. And remember to ask what support is included as standard with the system, and how the supplier will guarantee that your installation will work.
Mistake 3: Don't assume obstacle avoidance suits every application
A highly flexible AMR or AGV that can move around any obstacle dynamically may seem like a no-brainer, but such obstacle-avoidance behavior does not suit every application.
When it comes to navigating obstacles, there are two navigation approaches: virtual path following, whereby a vehicle stops and waits in case of a blockage, or obstacle avoidance, where a vehicle, usually a smaller mobile robot, dynamically adapts its path to move around the obstacle before returning to its designated virtual path.
Which approach is more efficient? The answer depends on the application in question.
Robots used in commercial-cleaning applications, for example, usually adopt an obstacle-avoidance approach to complete their missions. These tend to be light vehicles, operating at low speeds in highly congested environments, full of people, and the order in which they cover the area they are cleaning is not so important.
For many industrial applications however, virtual path following is often, counterintuitively, the most efficient approach. AGVs and even some AMRs can be large and heavy, and space might be limited, making obstacle avoidance both unpredictable and potentially more time-consuming, as a large vehicle tries to find an aisle space wide enough to navigate its way back to its target path. Not to mention that the idea of large vehicles moving around independently might add risk, or at least perceived risk, for existing human workers.
For one tire manufacturer in North America, a change from 37 robotic vehicles that used obstacle avoidance to 30 vehicles that run on natural-feature navigation configured for virtual path following, resulted in a total productivity increase of 27%.
Be careful to consider which navigation mode will work best for your application and take note also that some vehicles are available with a choice of either obstacle avoidance or path following.
Mistake 4: Not planning for a flexible future
Think about how your automated fleet might need to evolve as your business changes and grows. Do not assume that a single vehicle model, or single AMR/AGV manufacturer, will be able to meet all your needs in the future.
If you bear in mind the possible future scale-up of your fleet, think about how an AMR or AGV today might integrate with other types and brands of vehicle in the future. Does your current preferred vehicle supplier offer all the models your business might reasonably need? There are obviously risks involved in being locked into a single supplier, and a relatively limited product roadmap is one of them.
Your ideal AMR or AGV solution will be one that is inter-compatible with different brands. This will give you the widest possible choice of vehicles with which to build out the best-fit fleet for your organization.
David Béguin is sales director—North America at BlueBotics. Contact him at [email protected].
