Presented by:

DHL Looks to Collaborative Piece Picking for Flexibility, Optimizing Workforce

May 21, 2018      

The holy grail for supply chain automation is mobile picking robots, according to many in the industry. However, no single solution has emerged yet, and e-commerce and shipping companies are looking at multiple routes to increased efficiency. DHL is evaluating collaborative piece picking as a way to take advantage of human and robot capabilities for small-batch, high-variability orders.

The global market for collaborative robots was $112 million in 2015 and is growing rapidly, notes Grand View Research. Cobots with a payload capacity of up to 5 kg (11 lb.), suitable for pick-and-place operations, accounted for more than 40% of that market, says the analyst firm.

In 2016, the global warehouse robotics market was worth $2.28 billion, and it will experience a compound annual growth rate of 11.8% between 2017 and 2022, predicts Markets and Markets. Collaborative piece picking robots combine features of cobots and mobile robots in warehouses and e-commerce fulfillment facilities.

Robotics Business Review recently spoke about collaborative piece picking with Adrian Kumar, vice president, solutions design North America, at DHL. At our Robotics & AI Summit at LiveWorx next month, he will be presenting a “Case Study: Robotic Piece Picking Success Stories.” Here’s a preview of Kumar’s session:

DHL's Adrian Kumar will be speaking about collaborative piece picking robots at LiveWorx.

DHL’s Adrian Kumar will be speaking at the Robotics & AI Summit at LiveWorx ’18.

What will you discuss at our Robotics & AI Summit?

Kumar: We’ll talk about collaborative piece picking. From an engineering perspective, we’ll compare and contrast different methods in the marketplace. They range from simple, like replacing a cart, to more complex strategies, like swarming robots.

We’re unbiased and technology-agnostic. DHL is just looking at our different customer profiles to understand the best applications from their perspectives.

What’s driving the push to mobile collaborative piece picking?

Kumar: Traditional picking methods can be inefficient, as e-fulfillment requires lower density, or fewer items per order, and more items in the warehouse. Orders get harder and harder to pick.

Especially with labor shortages and the need for quick turnaround, there are high expectations. It can be hard to pick with multiple carts.

Traditional automation options required a lot of hardware — pick towers, loading goods-to-person solutions. There’s a pretty high cost of entry.

With collaborative piece-picking robots, there’s an opportunity to introduce automation into the current pick environment, not disrupt or rearrange the setup, and see significant productivity gains.

Where can the next generation of robots fit in?

Kumar: There are lots of opportunities to install robots in a modular fashion — [an operation] can start with six, and end up with 50 robots that have been deployed.

Robots can help human workers by reducing the amount of travel and becoming a pick-assist device. The human gives instructions, and the robot stops right in front of [the items to be picked]. Robots are bringing pickers to the next level.

Historical automation is still relevant for companies that are sure of their market and have stable types of goods. Collaborative piece picking provides additional flexibility.


Is Your Warehouse Ready for Robots?

Learn how robots are transforming warehouses. This download will help you evaluate your existing warehouse operations and prepare to successfully deploy logistics robots.
Download Your Free Guide

Are there trade-offs between types of robots or between robots and humans?

Kumar: We find that when you put these robots in the right operations, there aren’t that many trade-offs to worry about. There are not the traditional trade-offs of capital versus labor.

Sometimes, companies are getting creative with leasing, taking some of the capital investment costs out of the equation. There are some additional systems-integration costs.

If you want to get as productive as possible, you might put in a system that requires more capital. With 10 robots to one human picker, you’re looking at more capital, systems integration. You’re changing how orders are done by people to batch picking.

If you want to go in very light, you can just go with a “follow-me” robot of a 1-to-1 robot-to-person ration. It’s not as integrated, but you no longer have to push a cart or go to picking station.

How is DHL responding to workforce limitations with automation?

Kumar: We’ve had a hard time finding people for our omni-channel operations as we take on more product line. We want to merge different divisions into our operations, as we send e-fulfillment orders that others were processing before.

We couldn’t find the right people off the street, and robots help with the increase in volume from e-commerce. There is some flexibility with a 10-1 robot-person ratio. For productivity, you can adjust that ratio or make a big investment in infrastructure.

It’s no longer just a goods-to-person solution like Kiva [Systems, now Amazon Robotics]. We like the ability to handle average to heavy, peak demand. Locus Robotics or 6 River Systems can still supplement picking with robots and traditional cart picking, and we can also throw people at it.

How did DHL’s recent picking challenge demonstrate the current state of collaborative piece picking?

Kumar: Our challenge in Germany was different from that of Amazon. For the past three years, it asked contestants to pick certain SKUs out of a bin with a whole bunch of items. Some required picking.

The first two years, Amazon asked contestants to pick the right items out of a Kiva bin brought to them. That was a very complex challenge.

Then in 2017, it changed the challenge. Amazon no longer asked contestants to pick out of a Kiva bin. The totes were presented to the robots directly, and the robots could be overhead, so there was no awkward reach. They still had to pick from some mixed items.

Our challenge was to imagine that you’re in a current warehouse environment with shelves. Could you have a robot pick items out of shelves?

Vecna was the winner, and we’ve just started talking with other companies. For instance, we looked at rack-climbing robots from Israel that could pick a case out of a rack.

This is the next level of where you’re automating the full process instead of doing the actual pick. The technology is in its infancy, and companies are dabbling in different approaches.

DHL has also worked with Fetch Robotics and Seegrid. Are there any clear leaders among robotics providers?

Kumar: We certainly see that there are different collaborative piece-picking methods: You could have a swarm of Locus robots or a 6 River Systems robot that leads a picker. Both approaches are valid.

We’re also interested in robots that do traditional pallet moves and forklift-type moves. We’re looking at AGVs [automated guided vehicles], pallet movers, and collaborative piece picking.

Companies like Kindred are making a foray into fashion, with the ability to pick clothing. Every unit in the bin needs to be picked — in downstream or upstream picking activity — giving it to the robot for final sorting. We’re keeping an eye on that as well.

There could well be an operation, depending on our needs and labor availability, where we could put in two or three technologies.

What are some other examples of supply chain automation that you’ve looked at?

Kumar: There is robotic storage like Swisslog’s AutoStore in a tight grid, where products are extracted with robots in totes. They go on a conveyor to a stationary picking robot, and then to another robot that performs the sort process.

If your goal is “lights-out” automation and people are not available, that’s what it might look like. Today, the economics aren’t really there for us to do anything like that.

Why is flexibility so important for collaborative piece picking?

Kumar: Labor availability is the key factor. For the most part, when you need some degree of flexibility, where items are not the same size and have varying profiles, you might design a system for small orders.

When you start automating to the Nth degree, you would design storage systems, robots, the order-sort bin to predetermined sizes — you can vary stations only slightly. As soon as something changes, what’s your recourse?

If it’s 10 lb. and sorting to the largest bin, and then you get larger orders, you have limited options if you want a lights-out system.

You could send these types of orders with robots and then have a manual process. Lights-out might not be desirable in all circumstances. You also don’t want to be very rigid in working with customers.

Is DHL looking for robotics partners, or would it buy a company to bring those capabilities in house?

Kumar: I don’t know if there’s an advantage to buying a robotics company. We have a lot to gain by tapping into the incredible creativity and entrepreneurship of all these companies, taking a best-in-class approach to what’s best for us and our customers.

We want to pick partners who have their act together and work with them on enhancements to their technology. We would make arrangements that could be exclusive to us.

Our digitization strategy is really important to us. We’re making investments, encouraging hackathons, trend communities, and startup labs. Our company is very committed and understands that a lot of this creativity comes from the individual. We’re also doing it internally.

“We know the world is changing, and it’s ripe for disruption. To remain competitive, we have to understand different solutions.”

Adrian Kumar, vice president, solutions design North America, DHL

How is the competitive landscape, with everyone considering adopting more automation?

Kumar: We know the world is changing, and it’s ripe for disruption. To remain competitive, we have to understand different solutions.

3PLs [third-party logistics providers] aren’t resting either; they’re looking at robotics. At trade shows, we see the competition stepping up its game.

Overall, if it’s done right, there could be huge growth. The amount of volume going through warehouses and the rise of e-fulfillment are tremendous.

There’s the Amazon effect, and we have different things happening within our industry.

Where does automation fit into last-mile delivery?

Kumar: For deliveries, crowdsourcing deliveries through Uber or Amazon Flex has had an impact on the business.

We’ll continue to research drones and autonomous vehicles, but the more immediate and relevant solution is the crowdsourced delivery angle.

We recently launched DHL Parcel Metro, which gets us back into the last-mile delivery space. We can tap into the on-demand marketplace — people driving in a direction, looking to make some extra money.

We have last-mile arrangements with the U.S. Postal Service, but for same-day delivery, we’ve got to get creative. There was the courier service in the past, but that was price-prohibitive.

Editor’s Note: DHL’s Adrian Kumar will be presenting a “Case Study: Robotic Piece Picking Success Stories” at 1:15 p.m. on Tuesday, June 19, 2018. Register here to attend the Robotics & AI Summit @ LiveWorx ’18 in Boston.