Will The Pandemic Accelerate Automation In Supply Chains?

From Reuters Events
By Alex Hadwick

Meeting e-commerce demand without crippling costs is a huge challenge for logistics but a fundamental shift in how we build, position and operate fulfillment systems might change all of that. Reuters Events: Supply Chain investigates

It’s no secret that e-commerce has been a challenge for a huge number of brands, a challenge that has grown since the advent of COVID-19. The tension is between competing in a way that matches consumers’ desires whilst maintaining margins, with even brands as huge and powerful as Nike and Amazon reporting or expecting quarterly losses even as e-commerce operations boom in 2020. This is because consumer’s desires to buy online and send deliveries to their homes have accelerated far beyond forecasts made for three or even five years in the future in the space of just six months

Organisations can’t just keep bulking up operations in a conventional way. There isn’t enough warehouse space near key markets. There aren’t enough available workers who can help when demand peaks. There isn’t the transportation capacity in the last mile. Critically, there isn’t the cash available to keep scaling up our current way of doing things given the expected trajectory of e-commerce.

The only real solution is a radical rethink that puts more automation into fulfillment centres and distributes those fulfillment operations closer to the consumer in smaller spaces than have been attempted before.

Macro change – micro-fulfillment

Put simply, the current mode of fulfillment is unsustainable for many. The costs of delivering goods to consumers were already burdensome for a host of retailers in 2019. Skip forward to 2020 and we are looking at a whole different ballpark for those expenses, particularly in the last mile.

The massive and sudden demand for goods to be delivered direct to consumer’s homes, as well as changes in consumption patterns has sent many scrambling for capacity and delivery capability. As we covered in part one of our series on automation, there is also the need to account for worker safety, reducing the capacity for many facilities. On top of this, warehouse space is at a premium right now, adding additional expense. It is therefore an incredibly tough environment where companies are trying to squeeze every ounce of productivity out of their facilities and workers.

The second part of this is the potent effect the pandemic has had on the physical shopping space. With many shopping streets around the world already feeling the strain before the arrival of COVID-19, the advent of social distancing and quarantine has broken the back of many retailers. Bankruptcies have been widespread and we are not yet at the end of the tough trading conditions that have brought on those liquidations. Suddenly there is a lot of real estate looking distinctly empty in and around city centres.

We are therefore at a crossroads where moving more fulfillment into smaller facilities closer to consumers is looking increasingly attractive.

“I think there will be smaller distribution centres, close to urban areas, which are really specialized in piece picking,” thinks Markus Schmidt, President of Swisslog Americas.

Fergal Glynn, VP of Marketing at 6 River Systems, agrees seeing “What previously were massive stores” that ringed cities getting turned “into fulfillment centres.

“We do see as well that outside of the big box stores, there could be retail stores that are even smaller, that will become fulfillment centres,” believes Glynn.  

Already Terrie O’Hanlon, CMO at GreyOrange, is seeing that those companies that “Were able to immediately flex into omnichannel and e-commerce fulfillment,” when stores were forced to close “were in better stead than others who had more of a kind of a firewall between those two sets of inventory. I would say that’s in the past.”

Automation at scale but scaled down

Whilst this seems like a neat solution that makes the most of the current crisis, there is a reason why this model is not yet widespread.

The logistical demands of e-commerce has traditionally meant that the only way to account for the costs put back onto the company for picking and delivering items, while still being able to deliver rapidly to the consumer, was to operate at scale.

That scale has largely been massive. Most warehouse specialising in e-commerce fulfillment try to take advantage of economies of scale and usually sit in the hundreds of thousands of feet of floorspace or even millions. Amazon, for example, has largely taken up leases for warehouses of around a million square feet in the recent past.

The reasons are economies of scale that result in a chicken and egg scenario where it is very hard to make operations smaller. You need to keep enough inventory on hand to meet consumer demands, which are unpredictable. Therefore, this usually means large volumes.

How do you then sort, pick, pack and distribute this volume? Largely this has been answered through large-scale automation, which is expensive and often requires a large floorspace, with pick and pack stations linked by conveyor belts, further incentivising huge facilities.

Then at the store level, “The biggest trick has always been how do you keep the inventory saved, or basically stored as efficiently as possible, but still let people walk through it and pick it themselves?” asks Lior Elazary, founder and CEO of inVia Robotics, presenting another problematic dimension to putting fulfillment into stores.

“The problem is that customers expect great selection with very, very low lead times. The only way to do that is to pre-deploy inventory. The only way to do that is to create automation solutions,” says Scott Gravelle, co-founder, CEO & CTO of Attabotics.

And it is this new wave of automation that is changing matters, as robotic aid has become smaller and more cost efficient, making a two-tier highly automated system at last functional and attractive.

Major, centralised warehouses will still in all likelihood be necessary but the most sought-after inventory can now be positioned closer to consumers through micro-fulfillment centres that reduce distance and last mile costs.

“The difference between the traditional sort of omnichannel fulfillment” and the new wave says GreyOrange’s O’Hanlon, “is that you have a bigger operation that’s accommodating multiple form factors of automation to both replenish stores and to answer e-commerce demand,” that then feed into smaller operations closer to the consumer.

This is why 6 River Systems are “Investigating everything a customer could need for fulfillment, whether it’s in a half million square foot building, all the way down to your local retail store,” says Glynn.

Making micro-fulfillment work

The challenge is not just in terms of scale but also in the nitty gritty of functionality of automation in these smaller spaces.

“You’ve got to have automation in those facilities, which is extremely reliable,” points out Schmidt. This, once again, comes down to economies of scale. “If you don’t, you cannot afford to have a lot of technicians running around. It’s too small. If you build a big distribution centre, you have so much throughput and the system is so big, you [can afford] to have technicians on staff, no problem. However, the small fulfillment centres are typically a bolt on to store” and this means there isn’t the room for specialisation to the same degree “because otherwise you need to employ additional expensive people, which defeats the purpose.”

So, we have issues around scaling, cost, reliability and size that need to be solved before we can really consider the idea of semi-automated micro-fulfillment centres possible.

One company that thinks it has cracked the code is Attabotics. The start-up from the heart of Canada in Calgary makes the astonishing claim that it can shrink the footprint of a warehouse down to “12% to 15% of its previous footprint.”

The secret lies in a rather surprising place: Leafcutter ants.

“I asked what would be ideal for a robot not for a person. Then I ended up ripping off leafcutter ants,” says Gravelle. “I’m really, really glad that they don’t have IP representation, because we’d be in trouble.”

The ideas is to take “Their three-dimensional distribution of goods and create a 3D robotic shuttle,” which would lead to “the highest density storage solution currently being offered, while also providing the additional functionality of sequencing and sortation buffering and conveyance all in one footprint.”

By going vertical and compressing storage through robots passing through a matrix of bins and eliminating conveyor belts, Attabotics thinks it has cracked the complexity behind micro-fulfillment.

Attabotics’ system first asks their robots to get totes with the various items requested as part of an e-commerce order and “brings those items to a single point…. One system can take those three items, put them in a box in a single touch in a single location.” That box is then “put back into our system and sorted by postal code or carrier or by time and then delivered from there into the truck or it can now be used for micro fulfillment, it can be delivered to a transport vehicle.”

Gravelle claims that a small order of “items can be picked packed and shipped within about 90 seconds of receiving the order, depending on the backlog.”

These micro-fulfillment systems are supported by “a larger warehouse that’s out of market that actually can backfill” items, “So we don’t have to carry all the inventory in a single place.”

The future of shopping

So, it seems we’ve got strong conditions for the segmenting of the delivery structure into major centres feeding smaller distribution nodes closer to the consumer, driven by economic, social and technological changes. Where does this leave the future of retail?

Elazary believes it will be a dramatic shift from the once dominant high street. “I think it will probably be like an 80-20% type of thing, where maybe 20% of the store is just showcase, and then 80% might be, fulfilling orders and things like that. Then they’re closer to the customer.”

O’Hanlon also thinks we should reimagine the store to “More almost like a vending machine in a way, where … the person walks … up to the kiosk and they’ve been assigned online to pick up their order, or they can walk directly up to the kiosk themselves and place an order there.”

O’Hanlon gives the example of a shoe store, where “The person can actually observe the robots going to get the shoes, almost as an entertainment factor. The shoes come and you they’re delivered to a locker or a similar means and you can try them on. Then at that kiosk, you can say, ‘Now give me a half size bigger’. So, there’s a whole shopping experience … and the robots are taking any inventory that you don’t take with you and obviously immediately putting it back into stock.”

This will only require “Three or four people in the back of the store who are receiving inventory and putting it into the mobile stock units,” and providing customer service and this store will then also be able to fulfill e-commerce through the same systemic approach.

The next step for refining the micro-fulfillment approach so that it can become this widespread will be the adoption of demand-predicting machine learning. This is critical, as the interplay between inventory held in larger, more remote warehouses needs to be kept in a careful ballet of effective replenishment between there and the micro-fulfillment frontline.  

“The biggest cost with most retailers is the inventory they’re carrying, which is just a buffer,” notes Gravelle. “So, if you can create more transparency in optimisation, you can lessen the size of the buffer, thereby freeing up more of their capital for digital transformation, or for recovery, or for expansion or, ideally, for paying their staff more.”

According to Gravelle this means “Monitoring all the influencers of purchases, and I don’t just mean Kim Kardashian, but what she does tweet changes demand. I mean, the weather, holiday schedules, major sporting events, the latest rap video, all of these things can influence what gets sold in markets.” Using IoT and big data to correlate the influential factors and eventual sales will mean predictive analytics can place inventory where it is really needed and eventually be expanded to the transportation network as well.

“Most companies have historically looked at software for this piece, but at the end of the day supply chain is about moving atoms. Rethinking how you move atoms but also utilizing all the emerging technology” is going to be the key that unlocks this last mile segment of the distribution puzzle.

Read the full article at Reuters Events.