Photo: NRF
NRF 2020 Review: Human vs. Machine
Two of the biggest takeaways from the National Retail Federation Big Show in New York this week were at either end of the human-robot spectrum. Sure, there was plenty of talk about how retailers are taking advantage of AI and machine learning, how they are creating new ways to engage shoppers at the checkout and what they’re doing to bolster their sustainability efforts, but the real innovation is going on with the consumer-facing activities of both humans and machines.
For human labor, there was plenty of information about workforce management, task management and gamification. And there was a slew of new user interfaces to help push decision-making down to the front-line employee’s smartphone. All of these solutions are designed to help retail workers better meet the exact desires of shoppers in real time so the physical store can compete more effectively with digital-only merchants. A variety of vendor solutions have amped up the possibilities for management to communicate with store employees by moving from straightforward list-based instructions to more elaborate and interactive programs.
For robots, there were additional capabilities, business use cases and a somewhat clearer ROI in the near future. Retailers are increasing their investment in robotics technologies at the store and throughout the supply chain.
For both human-to-human systems and robotic ones, traditional retailers are pushing back against the digital tide with engaging new solutions. For example, Neiman-Marcus is using an innovative fitting room assistant app to help merchants better address the needs of the shopper on the spot with everything from accessories to shades and colors. And Walmart, as RetailWire reported earlier this week, is implementing a new e-commerce fulfillment DC that will have robots picking orders, a technology that could eventually be deployed at stores as well.
Exciting times!
Discussion Questions
DISCUSSION QUESTIONS: Where do you see the most promising intersection between humans (staff and customers) and robots in the next couple of years? What do you think the human-machine relationship will look like in 2030?
Poll
BrainTrust
Oliver Guy
Global Industry Architect, Microsoft Retail
Cate Trotter
Head of Trends, Insider Trends
My own opinion is that robots will be great working with products and other inanimate objects. So picking, washing floors, checking shelves for holes.
But in interacting with customers, I just think.it’s a bad idea. Why go to a store to talk to a robot? Just doesn’t make sense to me. I’d rather stay home if I have to talk to a robot.
The spectrum involving robots and humans is paralleled by the spectrum of low-touch to high-tough retail where the in-store team member is a critical component. Contextually, robots and other incredible technologies did not appear to be a primary concern for retailers (that we met with) of higher-touch categories – particularly when in relation to in-store operations. No doubt in lower-touch retail environments robots, machines, and other artificial human intelligence will continue to grow. “Amazon Go”- like stores will continue to open and be successful, and we will no doubt see some form of virtual help within these retail environments.
The lines and boundaries are clearly forming between the need for person-to-person connectivity and the requirements around efficiency, speed, accuracy, and supply chain cost optimization. From my perspective, it was very apparent that there were clear distinctions at NRF 2020 around what experiences bring you joy and how automation and robotics could save you time and money.
While the integration of AI, IoT, machine learning and automation into core retail functions such as merchandising and assortment planning were apparent, it’s the blending of the art (human creativity/ingenuity/interaction), and the sciences that will help retailers keep up with the changing consumer market.
It’s clear that robotics has its place with fulfillment centers, inventory management, etc, however, the reason we shop at stores is for the experience of connecting with the community, and engaging with an expert. Nobody wants to chat with a robot in aisle 2.
Engaging with an expert? In 10 years that robot may have a whole lot more expertise than anyone in the store and may be friendlier, too.
Perhaps you are right Gene, however, in 2020, were are not at that level of maturity with robots. I will hedge my bets on the human element, especially in the specialty and luxury retail segments. Nobody is going to buy a diamond ring from a robot yet
This is a hard one. If I think back 10 years and look at what we would have projected for today, I believe we (or at least I) would have been far off. We have moved much faster with implementation and use of technology than I imagined. I think the same will apply in the next 10 years.
Today I saw on the news a story of contact lenses that will do what Google glasses were designed to do and more. Yesterday, I would have thought such things were decades away.
In 10 years, we don’t even know what robots will look like. They likely will not be awkward or cute machines, but very human looking androids. If they are introduced widely in the next six or seven years, by 10 years from now we may feel very comfortable dealing with them. I am imagining checking in for an international flight. A robot could easily do what the representative does, and always be friendly about it.
I guess my net/net is that interaction will not be an issue in 10 years, but will be commonplace and comfortable.
Let’s first get it out of the way: what’s called “AI” isn’t intelligence but automated systems running on data. They’re not replacing the humanity of a person soon — but will brilliantly automate repetitive and boring tasks.
That said, the is deciding where human intelligence is most critical within your company. Without a focus on the human, a company sets itself up for competitive failure (at least once we’re past the first two or three years of the new tech hitting).
From what I see and the research I have done, the biggest place robotics and automation will take hold is in places where it drives speed, efficiency and eliminates cost. In the short term it is not necessarily going to drive sales – and it may well be that take up of anything focused that will need solid proof before adoption.
By 2030 I would hope that repetitive tasks are removed from humans to enable creativity (and customer-centricity) to flourish. Ease of teaching a robot (physical or otherwise) will be streamlined and these micro-engines of automation will be seen as key allies.
I read an article not too long ago about the life on the sales floor of an Old Navy store associate, which involves having to use seven apps to do their job. Working on the front line is getting harder.
NRF was definitely a tech show, I saw technologies to solve just about everything and I’m sure I missed almost as many as I saw. Robots are cool and can do amazing things, especially when they are doing tasks that free up humans to work with customers. I agree with Paula. I don’t go to stores to talk with a machine. I can be frustrated at home trying to get a usable answer from Alexa.
The most promising technologies positively improve shoppers’ in-store experience by enabling people–frontline workers and HQ–with trigger-based, data-derived, bottom-up insights into their business. We’re limiting our understanding of AI (and revealing antiquated thinking) when we limit the framework of the conversation to people vs. robots. I didn’t see many providers touting actual consumer-facing robots in stores (although certainly there were some).
Robots and AI will one day completely eliminate repetitive, mundane tasks for humans in retail… but not today! Yes, there are some immediate use cases that may work well for robots. Places like distribution centers and warehouses where robotic automation can make picking and packing processes more efficient, in grocery stores monitoring the aisles for spills (and cleaning them), performing shelf counts, taking temperatures in freezer cases (although IoT sensors may be better suited for that), and the list goes on based on demos at NRF. However, why are people going to stores today? Human interaction. Today, those robots will not provide a better, humanized, experience than a real store associate. Tomorrow? Once the robots have collected and processed many years’ worth of operational store data and customer interaction data? Perhaps. If at that point, customers find that the product knowledge those robots contain can help them make a purchase decision, then they may take the place of an associate. Until then, they can only provide an experience that’s much like the detached online experience consumers get today – so why would they go to the store then?
More than one study I’ve seen shows that there will be a net increase in human jobs over the next decade, as the machines take over some traditional human tasks. As the dust settles and more adoption of AI occurs, new human roles will emerge to not only make decisions based upon machine analysis outputs, but also jobs maintaining, repairing and operating the machines.
I think we need to level set our understanding of what a “robot” is. If we are thinking 1950s moving steel creatures, no, we probably don’t want them getting past the back room or warehouse just yet. Ditto if we are thinking industrial “dumb” robots that mindlessly perform the same function day in and day out. Unlike my good friend Ron, I don’t make a link between developments in AI and robotics. It’s a bit like the distinction between gasoline and a car. If you are going somewhere the car won’t get you there without gas — unless it is electric, of course — and the gas won’t do you much good if you don’t have a car to put it in. AI is the “gas” that powers the machine. So I see increasing uses of AI intersecting with humans in the form of smart devices, etc. at the front of the house and “robotics” being confined to the less public spaces. As someone who has been employed as a futurist for the past few decades, I’m hesitant to make predictions about 2030. As the old saying goes, “He who lives by a crystal ball soon learns to eat ground glass.”
Any discussion about robotics or machine intelligence in retail must consider the human factor. These are just tools, after all. They will add value where they empower or reduce the burden on human associates — especially with regard to dull, repetitive, physically difficult or dangerous tasks.
Customer-facing robots have yet to demonstrate a meaningful benefit to shoppers beyond trivial novelty. So far, I have not encountered even a vision of a future use that I consider compelling. Believe me, I searched for this at NRF without success. I have my doubts about this changing much by 2030.
The other major use case for something akin to robots in stores is shelf-scanning — those roving towers festooned with cameras that search for inventory voids and try to count items on hand. There were several examples on the exhibit floor. These are single-use roving camera systems with on-board image processing. They use a degree of AI to avoid collisions with people in the aisles, as they pursue their singular function. The same information may also be gathered using networks of less intrusive fixed sensing devices. It remains to be seen how soon the resultant data flows will reach sufficient accuracy to be operationally useful for activities like reordering and planogram compliance. In the mean time, humans with hand scanners remain the only viable option.
For these and other reasons, I see far more potential for robotics in the areas of movement and distribution of product and fulfillment of orders. Fulfillment centers and delivery networks seem to be areas of promise. “Hive” type warehouses and micro-fulfillment centers are gaining credence. Robots that pack orders for shipment will arrive in the coming decade.
So, behind-the-retail-curtain robotics will have an impact on shopper experience by enabling speedier and more accurate order fulfillment. They will eliminate many low-satisfaction manual tasks and create some pretty cool jobs for technicians. In the stores, however, humans will be the bigger story, as staff are better supported by AI tools and made freer to attend to shoppers.
Right now, robots can’t communicate at anywhere near the level that another human being can. There’s a place for them within some retail businesses, but it’s more in a service delivery role like behind-the-scenes product picking and so on.
In 10 years… who knows? Technology is advancing quickly and it may be that robots become human-like in their communication sophistication by then. The question is will customers want it? Will we want to talk to a robot in a store? Or would we prefer a silent robotic partner where we use our smart devices and apps to make requests and place orders that the robot fulfills? It’s important to remember that just because a technology can do something doesn’t mean that it matches up with the human expectation or want. That gap exists. It might narrow or even close in time but ultimately it’s about making the right choices of where to put technology and where to put human workers.
The sweet spot for robots is to do dangerous or repetitive tasks so humans are freed up to do what they’re really good at: creativity and engaging with other humans. That naturally leads robots to warehouse work (already prevalent), restocking, transporting, etc. Of course, humans need to design these systems. But also engage with people in-store, using AI and other digital support to build and grow authentic customer relationships.
There may be mechanical tasks that can be performed by robots, particularly as the US is not growing the workforce enough (either through immigration or birth). But there will be a series of smarter capabilities whose aim is to make humans smarter and more efficient, as ML based support capabilities. Customer facing is the main candidate for ML supported applications.