Consumers expect a wider range of products to be delivered quickly, and manufacturers and retailers are rethinking their supply chains. Fortunately, mobile robots are evolving to keep up with shifting e-commerce demands.

Robotics Business Review talked with Mike Oitzman, senior product-line manager for mobile robots at Adept Technology Inc. He noted that while robots can provide added value, there’s always room for innovation.

How is e-commerce affecting demand for logistics automation?

Oitzman: As I look at the market and talk to customers, I see the best practices of e-commerce emerging. With increasing demand for smaller order sizes and faster delivery, companies are evaluating the design and location of their warehouses and distribution centers (DCs).

In many cases, companies are looking at the size of their warehouses and DCs and the proximity to their target customers. Customers are making purchase decisions based on factors such as time to delivery and product availability.

Warehouses are now required to hold a larger volume of product, and they stock an increased number of types of products, especially during seasonal periods of high demand.

In addition, there is a trend to deliver perishable goods, which requires that the warehouse or DC be located relatively close to the consumer.

Automation is increasingly being tasked to fill the labor gaps that fall out of the decision of where to locate a warehouse. Some of the issues are trying to find skilled labor for a third shift or accommodating seasonal increases for promotions and the holiday buying season.

As people buy more online, orders are increasingly fulfilled 24 hours a day, 7 days a week, nonstop.

The size and location of a warehouse depends on many factors. Amazon, for example, has many regional mega-warehouses for its main product lines, but now that it has launched the Amazon pantry idea with fresh food, it requires a change in strategy.

Automation can help speed the entire logistics process — from getting incoming products stocked and shelved efficiently to the order pick and pack process.

The problems of handling a variety of products, handling fragile material without damage, and working in controlled environments all present new challenges for material handling automation.

How do mobile robots fit into modern manufacturing, warehousing, and logistics?

Oitzman: The predecessors to mobile robots have been around for a while in automated guided vehicles, or AGVs. AGVs are guided by magnetic tape or painted lines, and in many ways, they are not much different than a fixed conveyor.

A new generation of vehicles, led by Adept’s Lynx autonomous intelligent vehicles, or AIVs, is not constrained to marked aisles, thanks to natural feature navigation. These mobile robots can accommodate dynamic changes in the environment, such as people walking or driving forklifts throughout the building. They operate within exiting aisle ways without requiring building modifications for guidance.

This class of mobile robots is the first generation of collaboration technology finding its way into the factory and warehouse.

In addition to providing flexible material movement, mobile robots also provide traceability and predictability and improve safety for the workers required to work around fork trucks.

For example, the new Adept Lynx Cart Transporter enables flexible material movement between multiple steps of the manufacturing or logistics process, while simultaneously unburdening human workers from long-distance trips pushing carts in the facility. A positive side effect is that the human workers can be more productive at their workstations.

We’ve started to see mobile robots paired with grippers and arms. Is this a trend or just a particular use?

Oitzman: Mobile manipulation is still a technology hurdle, and a very complex problem to solve. We’ve got material movement down to a state-of-the-art logistics function of getting material from Point A to Point B.

The remaining problem is to automatically get product off shelves and into a shipping box. Material movement is the “low-hanging fruit,” as typically more than 25 percent of the time spent by warehouse workers is spent traveling around warehouses, while picking and putting is less than 10 percent of their daily tasks. The capability to see and pick accurately and quickly with a mobile robotic arm remains a difficult robotics research project.

We have deployed mobile robots equipped with robotic arms into some specific applications. Adept has successfully deployed mobile robots that load and unload machine tools in the semiconductor market. This application has a viable ROI [return on investment] because of the high cost of labor inside of the fab [chip-fabrication plants].

Our largest deployed fleet is made up of mobile robots with collaborative arms handling wafer pods; they are working next to humans in the fab. This is actually a simpler application than in the typical warehouse because the parts are more uniform and the operation is well-defined.

Our plan is to leverage this experience with mobile robots and collaborative manipulators to solve the needs of the logistics and warehouse markets.

In the fab, grippers are passive because everything we move is a standard size. Vision guidance is used to accurately locate the SMIF [standard mechanical interface] pods and the pick and placement locations.

As soon as you start trying to pick up a randomly placed shampoo bottle, a cereal box, etc., the arm motion isn’t the hardest part; vision guidance and accuracy is. It’s an exciting area — lots of people, including ourselves, are working to solve it in the next year or two.

What’s on your wish list of technologies to develop?

Oitzman: Mobile picking is on the road map. Adept is uniquely positioned to dominate in this market because of its long history on the fixed robot side. Once you combine a fixed, collaborative arm with a mobile robot, the next big problem to solve is vision guidance for the mobile system.

In addition, many of these sensors require additional processing power to deliver usable data in real time. This is an area that I believe needs to evolve quickly.

Note that in the recent Amazon Picking Challenge, all the vision processing was done off-board of the mobile robot.

We want to create a sensor and processing package in a size where it can operate effectively on a mobile robot, powered by a battery.

This sensor and processing package needs to be industrialized and robust, not a research model like many of the robotics startups coming to market today. Mobile robots with vision need to be easy to use, repair, and configure.

How will mobile robots in warehouses work with the Internet of Things (IoT)?

Oitzman: It’s a continuum — when I think about IoT, I think about small, smart, connected devices that I can access remotely through a standard, open API [application programming interface].

Whether it’s a bar-code scanner, a proximity sensor, a labeler, or even a palletizing work cell, making all of these devices accessible via the network means that each device can share its information or state with any other device.

This then makes it possible for the supervisory software to run the entire facility as a whole, rather than islands of automation throughout the facility.

Today, our Lynx Enterprise Manager software is responsible for managing a fleet of mobile robots operating in the warehouse or on a factory floor. The Enterprise Manager interfaces with the Warehouse Management System or the Manufacturing Execution System, taking a job and coordinating all of the mobile robot operations.

In a connected world, the Enterprise Manager will have access to all of the sensors in the environment and also be able to communicate effectively with the other automated equipment.

For example, cameras have evolved in the past 15 years. They once needed a separate computer to process the images, but many cameras can now do some or all of the basic image processing onboard, are Ethernet-connected, and can share that information with any other device on the network.

What are the market prospects for logistics automation worldwide?

Oitzman: We’re seeing growth in many places within the global economy. Traditionally, automation has been driven by developed countries because of a certain level of tech requirements and the availability of inexpensive labor in developing countries.

Now, however, we’re seeing the rapidly growing economies of Asia, specifically China, driving new requirements.

There’s a priority to reallocate labor to higher value-add jobs as manual production shifts to other less-constrained markets. This is driving innovation in robotics cost reduction and optimization, training, and ease of use.

Automation is now growing fastest in the places with the lowest levels of adoption. There’s some saturation in Japan, Germany, and the U.S., where many facilities are already automated, but others overseas are catching up quickly.

Last year, more robots were sold in China than in other countries, and we expect to see that continue.

Editor’s note: For more information on Adept’s new Lynx Cart Transporter, visit this page.