As 5G connectivity becomes more widespread—fueled by its great hype machine—the wonders promised seem incomprehensible at times. Remote control of fleets of vehicles from hundreds of miles away. Squads of cars or trucks that move in perfect synchronicity, communicating with not just one another but with the road itself. Shopping in a store where you never have to try on clothes thanks to augmented reality mirrors that do it for you. It’s easy to find breathless proclamations about 5G’s potential, but it’s difficult to find any research discussing what it will do for, or to, the average worker. Are we heading into a new golden age of productivity, a jobs apocalypse, or is much of the hype just marketing fluff designed to satisfy overzealous investors?
In Central Texas, locally adored grocery store chain H-E-B sits at the bleeding edge of retail technology, as near as a place to buy processed meats can be. The company’s response to the pandemic saw them leaning into their already successful curbside and delivery model, and in September 2020, news broke of their new partnership with warehouse automation provider Swisslog to deliver automation-heavy mini fulfillment centers (MFC) for standalone curbside pickup and delivery operations.
These centers will use a product called AutoStore, which is primarily a gridded, modular cube made up of vertical storage bins that are traversed and picked over from the top by wheeled robots. Doing away with the normal aisles required for human workers with heavy machinery like forklifts to pull product means the footprint of this system is drastically reduced. Instead, human workers are positioned at picking stations, where bins are delivered to them, containing, in sequence, the product for given orders. Efficiency gains here mean far fewer workers are necessary to perform the tasks of gathering product and packing orders, while, according to this promotional video, the robots “will never have a sick day.” That may sound appealing to some, but menacing to others.
H-E-B isn’t leading the way when it comes to automation. Colman Roche, Swisslog’s vice president of sales and consulting, said many retailers were able to easily hire workers for their warehouses after layoffs caused by the recession of 2008-2009 sent people scrambling for work. However, he said, this labor pool shrank over time as people returned to their previous careers during the ensuing recovery, leaving many open jobs and no available bodies to fill them.
Even smaller rural grocers seeking to compete with larger chains have internalized the experience of this rapid gain and subsequent gradual loss of available labor. As the covid-19 pandemic has had many of the same effects on the labor pool, some businesses, Roche said, are looking to MFCs—which can scale up capacity without a commensurate increase in workers—as a means to head off another dearth of workers, which he sees as inevitable, and likely to occur faster this time. This is the automated future that’s here now, but what happens when we put 5G in the mix?
I asked Roche how he expected 5G might affect jobs in an automated environment like AutoStore—which jobs might be preserved or taken away—and his answer was somewhat complicated. Currently, warehouse management systems are designed to be used by humans using things like barcode scanners, but are not well-suited for interacting with self-sufficient machines, which require much faster, higher-density transfer of information. A warehouse management system is built with the expectation of long pauses between transferral of data over a network. Automating that process means those transfers happen much faster, and information can easily be lost by systems designed for human interaction. Data packets can collide or fail to arrive.
Because of this issue, he said, automation takes place on isolated networks where there is no other traffic, so machines can communicate freely, reducing the likelihood of commands being received out of sequence or not at all. It’s currently not clear whether 5G will be able to handle both kinds of traffic simultaneously and whether that will make automation especially more likely to expand. The lack of real-world data on this, he said, will contribute to a hesitancy by warehouse operators to expand automation, and we should expect to see an adoption curve as it’s taken up. (Indeed, though enabling what is known as massive Machine-Type Communication is considered a primary use case of 5G by the 3rd Generation Partnership Project, ongoing studies are needed to confirm this capability.) Basically, 5G isn’t necessarily the silver bullet that will fully enable the 100% automation of warehouses that large companies and specifically Amazon believe we will see in the next few years.
In other parts of the retail and logistics world, 5G could change the landscape significantly. This could mean Minority Report-style changes like targeted in-store advertising, sure, but could also mean an obviously technology-driven experience, like virtual dressing rooms where customers try out clothes with AR via high-definition screens or rapidly cycle through different cosmetics products without physically touching a lipstick shade, for example. Though many of these changes are already being experimented with, the high-bandwidth, low-latency advantages of 5G can greatly expand AR capabilities within a store. For employees, this may make helping customers find things easier, or create less busy-work vis à vis returning clothing or other items to racks and shelves.
Though 5G could benefit workers’ efficiency or eliminate certain tedious steps needed to complete a task, it also endangers the need for in-person, human workers in retail settings. The labor dystopia that has been creeping ever closer to reality for decades might already be here.
Thanks to the potential ability to stream 8K video with imperceptible lag, video-based remote work could be a real possibility in retail and beyond. Lindsay Notwell of wireless WAN edge solutions company Cradlepoint, in an interview with Diginomica, said she envisions a future in which retail stores contract workers from a call center of “subject matter experts” who can work for many stores simultaneously. These workers would be accessible from camera-equipped kiosks positioned throughout stores, ready for on-demand interactions with customers. The article proposes that this will solve the issue of being unable to hire knowledgeable experts within the local labor pool, while also providing cost savings and a competitive edge to businesses that make use of the technology.
This is the sort of thing that makes a good pitch without context in a board room, but for retail employees, this could eliminate some of the few pleasures of working in a retail setting, such as a relatively free-roaming environment, work that keeps you somewhat physically active, or downtime in which you can often take care of personal business. Instead, it replaces them with assigned seating, more verbal assault from customers, and job security under constant threat from call metrics.
In logistics, a survey of machine learning researchers published in 2018 concluded that truck drivers could be outperformed by high-level machine intelligence by 2027. (Many skeptics decry full human obsolescence in trucking as infeasible, if not downright impossible, in the near term.) Other predictions have the trucking industry moving to a remote work model where trucks run constantly while the drivers control them in shifts from an office and go home to their families at night, while the trucks roll on. Like retail remote work, this could potentially be possible due to near-zero latency, high-bandwidth connections that allow for multiple 8K video streams transmitted from the truck to the driver.
There are even some far-flung ideas about trucks deploying drones to deliver from the trailer to the customer, eliminating the need for transfer of freight to local delivery in a warehouse. Many of these changes would rely on millimeter-wave 5G—the version of 5G that provides the incredibly fast speeds and almost-zero latency most often talked about in these contexts. Of course, true mmWave 5G will spread in urban areas, but will almost certainly not find as much footing in the rural countryside that makes up most of the United States, geographically-speaking, due to the need for a density of cell towers—mmWave 5G operates on high spectrum with short-range signals that can’t penetrate buildings, which presents another challenge.
More realistic, immediate possibilities of benefits for drivers and those they share the road with would come in the form of IoT sensors that predict mechanical failure, keep a constant monitor on tire pressure and nearby cars, and relay traffic conditions and possible re-routes in an instant, much more granular way than current GPS-mapping on a smartphone. These things, if realized as imagined, would contribute not just to drivers’ abilities to perform their duties efficiently, but would theoretically lead to increased safety as drivers are directed away from hazardous roadways or alerted of mechanical safety issues before they become problematic.
In many ways, existing writing from some industry publications and corporate PR about 5G’s potential seems so much like the singularity talk that’s been the domain of futurists for so long. And like Zeno’s Dichotomy paradox, it can feel like we are forever only getting a fraction, but never all, of the way there. Whether and when this potential is realized will depend on the planning and follow-through of those involved: the wireless service providers, governmental bodies both local and national, as well as the businesses that will use the products that 5G will enable. The advancement toward that future is not going to be tidy, but uneven, jerky. This intrinsically disorganized process even has a name: the 5G go slow cycle, which is shorthand for the reality that nobody owns 5G, and its rollout is entirely dependent on the operators that deploy it.
As for its effects on workers, this future is unclear, relying on the intentions of business to protect or listen to workers’ input and the appetite of governments to regulate the technology.
Christian Sweeney, organizing deputy director of the American Federation of Labor and Congress of Industrial Organizations (AFL-CIO) and Amanda Ballantyne, director of the AFL-CIO’s Technology Institute, said via email that they “have faith that technological change can accrue to everyone’s benefit but that only happens if workers’ interests are taken into account.”
“Will the nature of work change? Yes. Does it spell doom? Not necessarily if workers have a seat at the table.”
Unions can help protect workers whose jobs could disappear.
“When the collective bargaining process is working, workers can negotiate advance notice, opportunities for reskilling, and a voice in the ways that the companies are implementing technology,” the AFL-CIO reps added. “Additionally, public policy needs to include strategies to ensure that training opportunities, good jobs are created in place of those lost.”
In the U.S., despite four decades of aggressive union-busting on both sides of the aisle, cautious optimism is warranted. Encouraging signs that workers’ voices are receiving renewed priority include President Biden’s warning to Amazon against intimidation of organizing workers and his subsequent statement encouraging the House to pass the Protecting the Right to Organize Act of 2021 (it did), a bill that purports to strengthen unions through a variety of changes aimed at weakening ongoing efforts—right to work laws, mandatory company meetings designed to lobby against organizing, for example—to undermine unions. Provided the companies bargain the effects of the changes with their employees, some of the worst predictions of 5G doomsayers may never materialize.