Robots Don’t Get Sick

September 15, 2020

COVID and other current crises have revealed the value of work automation, with justifiable celebration.*

Scientists at the University of Liverpool have a new lab assistant with a very strong work ethic: a robot chemist that conducts experiments by itself. The 1.75-metre-tall intelligent robot moves around the laboratory, avoiding human co-workers and obstacles while performing a wide range of different tasks independently.  It can even decide for itself which tests to do next based on previous results.  A cylindrical robot rolls into a treatment room to allow health care workers to remotely take temperatures and measure blood pressure and oxygen saturation from patients hooked up to a ventilator. Another robot that looks like a pair of large fluorescent lights rotated vertically travels throughout a hospital disinfecting with ultraviolet light. Meanwhile a cart-like robot brings food to people quarantined in a 16-story hotel. Outside, quadcopter drones ferry test samples to laboratories and watch for violations of stay-at-home restrictions.

Should work automation replace people?  This is a debate as old as the dawn of technology, famously illustrated by the Luddites who opposed automated textile manufacturing machines.   As the COVID crisis accelerates work automation, crisis-driven imperatives may lead to short-term decisions, such as replacing humans with automation.  Costs may go down, risk may be reduced, and patients and customers may enjoy the novelty … in the short run.

Are such solutions sustainable and correct for the longer-run?

The paradox is that as the COVID crisis accelerates our experience and acceptance of automation, it also offers important opportunities to see beyond solutions like human-replacement automation, and instead to build a more complete and nuanced approach.

1. What are the new and important lessons being learned and experienced now, due to the crisis?

Lesson #1:  Work Automation Accelerates the Temptation to Replace Humans

The World Economic Forum notes two dozen ways robots have been applied to the COVID-19 pandemic, from health care in and out of hospitals, automation of testing, supporting public safety and public works, to continuing daily work and life.  A Time article quotes Daniel Susskind, a fellow in economics at Balliol College, University of Oxford, and the author of A World Without Work: Technology, Automation and How We Should Respond: “This pandemic has created a very strong incentive to automate the work of human beings … Machines don’t fall ill, they don’t need to isolate to protect peers, they don’t need to take time off work.”  Paradoxically, before COVID work automation was often seen as protecting human workers, by taking on dirty or dangerous work, but in the COVID era, automation is protecting shoppers, patients and others from potentially dangerous contact with human workers.

Tasks that were typically regarded as “human,” and safe from automation, are now being automated.  UCLA’s Mattel Children’s Hospital employs a Pixar-like robot named “Robin” to provide emotional support to young pediatric patients: “Robin’s technology enables the robot to build what is called associative memory — it recognizes a child’s emotions by interpreting his or her facial expressions and builds responsive dialogue by replicating patterns formed from previous experiences.” A bright-yellow robot, named “Fluffy,” has five “eyes” that capture images of an auto factory floor, helping engineers design upgraded workspaces.

Lesson #2:  Customers, Patients and Stakeholders Increasingly Accept Work Automation

As we become adept at virtual meetings, some companies offer “work from home forever.” Patients, grocery shoppers and others become accustomed to new technologies.  MIT Technology Review notes that Tally, an in-store robot checking shelf inventory, was “a little strange to shoppers” a year ago, but is now, mid-pandemic, “not even close to the most unusual thing happening inside the store … posing far less threat than other shoppers.”   A survey of 1,000 U.S. adults in late-April, 2020, suggested that COVID may be accelerating public acceptance of automation, AI and the collection and use of personal information:  Thirty-three percent said they are now more comfortable with robots in grocery stores; 30% said they were more willing to share anonymized personal information if it will make their community safer or healthier; and 33% said that the COVID era has made them more comfortable with the idea of robots helping doctors and nurses.  Paul Nunes, Annette Rippert and Larry Downes succinctly warn “don’t kid yourself about consumer resistance” to new online, digital and streaming alternatives.

Lesson #3:  Optimal Solutions Combine Not Replace Humans With Automation

Accompanying many of these stories are photographs and videos of smiling doctors, nurses, caregivers, factory operators, retail associates, and engineers.  It’s understandable and justifiable to celebrate how quickly human workers have embraced and learned to work with automation. HR leaders are integral in the operational support for these notable cases of accelerated work automation.

Yet, for each celebratory example, there’s the specter of automation replacing humans.   Once COVID subsides, will hospitals and patients want a return to humans providing patient comforting?  Should hotel food delivery and ventilator-patient vital sign monitoring return to their previous human work mode?

As Ravin Jesuthasan and I  emphasized:  Optimizing the value of work automation is seldom achieved through replacing human jobs with robots, chatbots or AI. Rather, it requires a more nuanced approach that combines humans and automation.  Each of the examples offers a range of human-automation combinations.  Optimal work automation sometimes replaces human work on a particular task, but human-automation combinations can also augment humans or even reinvent human work in ways not possible without automation.

Every example above offers choices, hinted at in the popular press.  Robots that can comfort children in the hospital might replace human nurses or counselors, but the technology might also “create new jobs” or “free up the nurses and counselors for more important work.”  Will robots replace, augment or reinvent the work of the humans?  The answer requires getting beyond broad hints, and capitalizing on the lessons available during the crisis, as a platform for sustainable lessons beyond the crisis.

2. Which of those lessons should sustain after the crisis?

Lesson #2 may be well baked-in the longer the crisis lasts.  Lesson #1 is also increasingly baked-in, and may actually result in a post-crisis tendency to snap back too far in the direction of human replacement.  Lesson #3 will be pivotal for leaders to sustain post-crisis.  Beyond the organization, the current crises have renewed attention and increased organizational accountability for institutional and social inequality.  It appears that a “replace the humans” approach to automation also risks exacerbating such inequality.

In “Robots and Jobs: Evidence from U.S. Labor Markets,” MIT economists Daron Acemoglu and Pascual Restrepo found that between 1990 to 2007, in the U.S., adding one additional robot per 1,000 workers reduced the national employment-to-population ratio by about 0.2 percent, with some regions affected far more than others. MIT News noted that each additional manufacturing robot replaced about 3.3 workers, and lowered wages by roughly 0.4 percent.  These economists also found that “A lot of the new job opportunities that technology brought from the 1960s to the 1980s benefitted low-skill workers … But from the 1980s, and especially in the 1990s and 2000s, there’s a double whammy for low-skill workers: They’re hurt by displacement, and the new tasks that are coming, are coming slower and benefitting high-skill workers.”

The key to sustaining Lesson #3 will be the values, frameworks and decision rules used to implement work automation.  Those are being shaped now, during the crisis, as organizations respond to the unprecedented acceleration of automation and its consequences.

3. Which sustainable lessons will be challenging, due to inertia, ignorance, or other factors that push to snap back to before … or worse?

Recall Lesson #1 above.  There will be significant pressure to replace humans with automation, as a rapid way to recover lost profits, reduce risk and meet the needs of more automation-accepting customers (Lesson #2).

For example, Blue Prism touts cost savings through reductions in headcount, payroll costs and benefits. Their client, Siemens’ saw the first 170 RPA Blue Prism bots take up 280,000 hours of human equivalent work.  The 80,000 bots run by the Bank of Columbia reportedly freed 127K hours of human work.  The Guardian notes that Capita, a FTSE 100-listed firm, said it removed 2,000 jobs as part of a cost-cutting drive, and would use the saved money to fund investment in automated technology. The Apple and Samsung supplier Foxconn was reported to have replaced 60,000 workers with robots, while the former chief executive of McDonald’s suggested a similar tactic in response to low-paid workers’ demands for better pay and conditions.  Wharton describes Momentum Machines, that built a hamburger-making robot that can reportedly produce up to 400 hamburgers per hour. “Our device isn’t meant to make employees more efficient,” co-founder Alexandros Vardakostas has said. “It’s meant to completely obviate them.”  Yet, the company also told Business Insider in 2012 that letting robots fill in for humans in the kitchen may actually promote job growth because automation would allow hiring new employees to improve their technology and staff additional restaurant locations.  (Thanks to Sue Cantrell for the information in this paragraph)

Consider the robots that comfort child hospital patients. One outcome might be that robots replace nurses/counselors.  Yet, another is that robots could do the basic care of the relatively calm children, and augment the work of human nurses and counselors, alerting them to situations where children need more than the comforting capability of the robot, thus bringing the human nurses and counselors to precisely where they are needed.  Finally, robot comforters might reinvent the work of nurses and counselors, creating new work where humans train the robot to become even better at reading emotions; or where human nurses/counselors use the robot and it’s “voice” as a portal to connect remotely with the most challenging cases, allowing the best nurses/counselors to reach children more quickly or at a greater distance.

4. For the challenging lessons, what are the pivotal and essential actions to take now, while the crisis provides motivation, attention and awareness, to avoid missing the window for change?

Work automation is nuanced, complex and uncertain, so leaders always have choices about its implementation.  “It’s not all doom and gloom,” says Acemoglu. “There is nothing that says technology is all bad for workers. It is the choice we make about the direction to develop technology that is critical.”

What lessons can be learned now and sustained beyond the crisis?

Work Deconstruction and Reinvention.  As crisis-driven work changes accelerate, confining “work” into a “job” and “worker” into a “job holder” forces a perspective that is incapable of revealing and optimizing the variety of ways to combine human and automated work.  Work operating systems will increasingly need to perpetually deconstruct jobs and workers into more granular units such as tasks and skills/capabilities.  This is vividly apparent during the crisis.  Robots or chatbots seldom replace all the tasks in a job, offering examples where automation might take on more repetitive tasks such as scanning shelf inventory, but not other tasks such as helping store customers or encouraging mask-wearing.  As the crisis makes work more fluid, it is an opportunity for leaders to learn and adopt work systems now, that deconstruct and then reinvent work, and that will be needed to sustain work fluidity after the crisis.  Such frameworks are described in Lead the Work, Reinventing Jobs, and by the World Economic Forum’s Future of Jobs Report 2018.

Work Design Beyond “Replace Humans.”  The current racial/social injustice crises have brought new attention to the values and decision frameworks of organizations, and their role in workforce health and equity.  Worker replacement can exacerbate inequities, so it is now that leaders can consider alternatives to human replacement with work automation.  Automation can also augment and reinvent work.  COVID has revealed countless examples of factories being retooled from making perfume, spirits, T-shirts and cars, to making supplies to fight the virus.  That involves human workers in creative combinations with automation.  These examples offer leaders important lessons to look beyond immediate human replacement, and consider questions such as “When should we slow down our automation plans to offer our human workers that chance to reinvent their capabilities, rather than replace them?”

 Automation As A Collaborator.  The COVID crisis has revealed how Zoom and other remote platforms remake collaboration, combining human workers and automation in new ways.  Machine-learning algorithms collaborate with scientists in a National Institutes of Health initiative to rapidly scan medical images of infected lungs and hearts. Just as songwriter Taryn Southern sees AI as her collaborator, the current crisis offers lessons about adding automation to the “social network.”

Engagement with Work Automation.  The current crises have brought renewed attention to the importance of worker engagement, and engagement with work automation is an important element.  Alongside skill enhancement, it will be vital to understand worker engagement, trust and motivation to support work automation.

Cross-Discipline Collaboration and Frameworks.  These lessons will be sustained not by the isolated efforts of HR, or any single organization function.  The current crises offer renewed urgency, and the chance for HR to work with the Technology, Operations, Information and other functions.  The sustainable lessons from the crisis will best be supported when all these disciplines bring the best evidence-based frameworks to optimize work automation, beyond worker replacement.

*   Deep thanks to Nora O. Hilton for her invaluable help with this article