Supply chain and labor shortage issues haven’t let up. In fact, nearly 11 million jobs remain vacant in the U.S. alone. As more retailers and e-commerce businesses look for solutions, the most innovative are turning to mobile robots to efficiently move, sort and package their goods.

If you’re considering implementing mobile robotics, keep in mind these four essential “T’s” as you design and execute your strategy.

Time to install. One important element to consider is how long it will take to install a system in your distribution center. When reviewing your automation options, look for solutions with shorter install times; if the installation takes several months, you won’t reap the benefits (e.g. ease of use, higher accuracy and faster time to delivery) until that’s been fully implemented.

Additional considerations impacted by install time include lost warehouse space while the systems are being installed, as well as lost systems that must be temporarily shut down in order to complete the installation. Combined, that can make a huge impact on throughput. 

One option to consider when looking to shorten installation times is modular designs. With these systems, some installations can be done in sections in just a few weeks. Installing in sections is especially beneficial for brownfield installs, i.e. when the installation of new hardware or software must temporarily coexist with the legacy systems they are replacing. In a brownfield environment, if you can replace 50% of existing throughput while occupying 25% of the allocated space, you minimize the impact on the remaining 75% since you’ve already delivered 50% of the throughput requirement. 

Time to transition. Another important “T” to consider is the time to transition. 

First, there’s worker training and onboarding. Any time you put in automation, you’ve traditionally needed engineers to run it — and good ones are scarce. Thankfully, new “consumerized” robotics interfaces (think iRobot’s Roomba) can make it easy for any warehouse worker to operate advanced robotic systems with minimal onboarding, removing the need for specialized education or rigorous training. 

These systems become even more user-friendly when paired with predictive alerts. By informing operators of conditions that will lead to throughput issues before they become an active problem, operators are able to proactively address and repair systems and keep them running smoothly without interruption or delays.

An additional element of time to transition is a support system. It’s important to have a strong support system in place that can provide assistance and parts (as needed) within a very short amount of time, as prompted by predictive alerts and scheduled maintenance. This is needed not only during initial setup, but also on a continuous basis after installation is completed so that repairs are done seamlessly with minimal downtime.     

We’ve seen these simplified user interfaces and comprehensive support systems help cut transition time from months to a few weeks, as well as reduce system downtime through issue prediction and advance resolution, so be sure to make these areas a priority when selecting which robotic automation to implement into a warehouse.

Throughput. Throughput is perhaps the most commonly considered “T” when looking into automation. However, even though it appears straightforward, it’s important to look beyond the traditional best practices/learned efficiencies approach. Instead, seek adaptive systems that get smarter by continuously learning throughput patterns of time and space.

Different goods flows change throughout company life cycles. Can your robotic automation system dynamically switch from one goods flow to any other goods flow? Can it reallocate systems to higher-need sections if a flow suddenly slows on one end and gets higher at the other?

This is where patterns of time and space come into practice. To optimize patterns in time, robotic automation systems must learn how goods flow over time as well as what may be needed in the future so that resources can be preemptively allocated for increased efficiency. For example, knowing that a certain flow of goods is higher than another early in the morning, or on particular days of the week, or specific to seasonality, allows a learning system to allocate resources in advance to match those flows. To optimize patterns in space, robotic automation systems must learn which sortation pathways have been most efficient in the past for particular goods flows so that those pathways are preferred in the future.

With the increasing use of AI and ML in warehouse robotics, dynamically optimized throughput is now possible as long as you ensure the chosen system implements the above capabilities. 

Total cost of ownership. Besides the initial upfront cost of a mobile robotics system, it’s paramount to consider costs incurred due to labor, transition time, maintenance and updates as throughput optimization, speed of training and ease of repair are vital determinants of total cost of ownership. 

As discussed earlier, modular designs are a great option to speed installation while improving throughput and reducing downtime from maintenance. Automation pods can be designed and configured from field-replaceable units (FRUs), with the units being swapped out later for maintenance, repair or replacement — with four-screw ease. By reducing lost time through installation optimization, the total cost of ownership can be drastically decreased.

Maintenance can be expensive both in terms of parts replacement and time loss. Most systems have a set schedule of maintenance: “After this much time, go replace X part.” But such schedules are based on expected usage, not real usage, leaving it up to you to determine frequency. 

Today’s more-advanced robotic systems will do predictive analysis on operations to give you a real schedule. They’re also intelligent enough to warn of pending problems and needed maintenance before they cause downtime. You can wait on replacements in some cases, saving money and improving total cost of ownership. Or if you’re running one part hard, you can replace it early so you don’t end up with a broken system waiting for the “typical” replacement window, greatly reducing the chance of unexpected lost profits.

In addition, rapid technology advancements are drastically speeding up the replacement cycles of warehouse systems. It used to be 20 years, but today it’s closer to 5 to 10 years. Therefore, it’s important to look for systems that are architecturally designed for slower obsolescence in order to reduce the total cost of ownership.

The idea is to keep your equipment constant while using software to make the system better and better. As orchestration learns from your operations, it automatically improves overall efficiency and continues growing along with you, slowing the rate of obsolescence as compared to non-learning systems.

Today, it’s no longer science fiction to see hundreds or thousands of mobile robots operating in a field together in close proximity and at high speeds. They’re in distribution centers, micro-fulfillment centers, even back-of-store operations. They’re enabling seamless curbside pickup and helping load trucks in a way that’s most efficient for deliveries.  

Supply chains are becoming increasingly intelligent and adaptive as a result, and those that don’t adopt and stay at the forefront of robotic automation are going to get left behind. Whether it’s your first time implementing robotic automation into your warehouse or you’re looking to update your current systems, keeping in mind the four “Ts” will ensure you’re getting the most out of your investment.

Sridhar Solur is chief product officer and general manager of mobile robotics at Berkshire Grey.