Rise of the Robots: Technology and the Threat of a Jobless Future (24 page)

Automated systems can also provide a viable second opinion. A very effective—but expensive—way to increase cancer detection rates is to have two radiologists read every mammogram image separately and then reach a consensus on any potential anomalies identified by either doctor. This “double reading” strategy results in significantly improved cancer detection and also dramatically reduces the number of patients who have to be recalled for further testing. A 2008 study published in the
New England Journal of Medicine
found that a machine can step into the role of the second doctor. When a radiologist is paired with a computer-aided detection system, the results are just as good as having two doctors separately interpret the images.
¹³

Pathology is another area where artificial intelligence is already encroaching. Each year, over a hundred million women throughout the world receive a Pap test to screen for cervical cancer. The test requires that cervical cells be deposited on a glass microscope slide and
then be examined by a technician or doctor for signs of malignancy. It’s a labor-intensive process that can cost up to $100 per test. Many diagnostic labs, however, are now turning to a powerful automated imaging system manufactured by BD, a New Jersey–based medical device company. In a 2011 series of articles about job automation for
Slate,
technology columnist Farhad Manjoo called the BD FocalPoint GS Imaging System “a marvel of medical engineering” whose “image-searching software rapidly scans slides in search of more than 100 visual signs of abnormal cells.” The system then “ranks the slides according to the likelihood they contain disease” and finally “identifies 10 areas on each slide for a human to scrutinize.”
¹⁴
The machine does a significantly better job of finding instances of cancer than human analysts alone, even as it roughly doubles the speed at which the tests can be processed.

Hospital and Pharmacy Robotics

The pharmacy at the University of California Medical Center in San Francisco prepares about 10,000 individual doses of medication every day, and yet a pharmacist never touches a pill or a medicine bottle. A massive automated system manages thousands of different drugs and handles everything from storing and retrieving bulk pharmaceutical supplies to dispensing and packaging individual tablets. A robotic arm continuously picks pills from an array of bins and places them in small plastic bags. Every dose goes into a separate bag and is labeled with a barcode that identifies both the medication and the patient who should receive it. The machine then arranges each patient’s daily meds in the order that they need to be taken and binds them together. Later, the nurse who administers the medication will scan the barcodes on both the dosage bag and the patient’s wrist band. If they don’t match, or if the medication is being given at the wrong time, an alarm sounds. Three other specialized robots automate the preparation of injectable medicines; one of these robots deals exclusively with highly toxic
chemotherapy drugs. The system virtually eliminates the possibility of human error by cutting humans almost entirely out of the loop.

UCSF’s $7 million automated system is just one of the more spectacular examples of the robotic transformation that’s unfolding in the pharmacy industry. Far less expensive robots, not much larger than a vending machine, are invading retail pharmacies located in drug and grocery stores. Pharmacists in the United States require extensive training (a four-year doctoral degree) and have to pass a challenging licensing exam. They are also well paid, earning about $117,000 on average in 2012. Yet, especially in retail settings, much of the work is fundamentally routine and repetitive, and the overriding concern is to avoid a potentially deadly mistake. In other words, much of what pharmacists do is almost ideally suited to automation.

Once a patient’s medication is ready to leave a hospital pharmacy, it’s increasingly likely that it will do so in the care of a delivery robot. Such machines already cruise the hallways in huge medical complexes delivering drugs, lab samples, patient meals, or fresh linens. The robots can navigate around obstacles and use elevators. In 2010, El Camino Hospital in Mountain View, California, leased nineteen delivery robots from Aethon, Inc., at an annual cost of about $350,000. According to one hospital administrator, paying people to do the same work would have cost over a million dollars per year.
¹⁵
In early 2013, General Electric announced plans to develop a mobile robot capable of locating, cleaning, sterilizing, and delivering the thousands of surgical tools used in operating rooms. The tools would be tagged with radio-frequency identification (RFID) locator chips, making it easy for the machine to find them.
¹⁶

Beyond the specific areas of pharmacy and hospital logistics and delivery, autonomous robots have so far made relatively few inroads. Surgical robots are in widespread use, but they are designed to extend the capabilities of surgeons, and robotic surgery actually costs more than traditional methods. There is some preliminary work being done on building more ambitious surgical robots; for example, the I-Sur project is an EU-backed consortium of European researchers
who are attempting to automate basic procedures like puncturing, cutting, and suturing.
¹⁷
Still, for the foreseeable future, it seems inconceivable that any patient would be allowed to undergo an invasive procedure without a doctor being present and ready to intervene, so even if such technology materializes, any cost savings would likely be marginal at best.

Elder-Care Robots

The populations of all advanced countries, as well as many developing nations, are aging rapidly. The United States is projected to have over 70 million senior citizens, making up about 19 percent of the population, by 2030. That’s up from just 12.4 percent in 2000.
¹⁸
In Japan, longevity combined with a low birth rate make the problem even more extreme; by 2025 fully a third of the population will be over sixty-five. The Japanese also have a nearly xenophobic aversion to the increased immigration that might help mitigate the problem. As a result, Japan already has at least 700,000 fewer elder-care workers than it needs—and the shortage is expected to become far more severe in the coming decades.
¹⁹

This surging global demographic imbalance is creating one of the greatest opportunities in the field of robotics: the development of affordable machines that can assist in caring for the elderly. The 2012 movie
Robot & Frank,
a comedy that tells the story of an elderly man and his robotic caretaker, offers a very hopeful take on the kind of progress we’re likely to see. The movie opens by announcing to the viewer that it is set in the “near future.” The robot then proceeds to exhibit extraordinary dexterity, carry out intelligent conversations, and generally act just like a person. At one point, a glass is knocked off a table, and the robot snatches it out of midair. That, I’m afraid, is not a “near future” scenario.

Indeed, the main problem with elder-care robots as they exist today is that they really don’t do a whole lot. Much of the initial progress has been with therapeutic pets like Paro, a robotic baby
seal that provides companionship (at a cost of up to $5,000). Other robots are able to lift and move elderly people, saving a great deal of wear and tear on human caretakers. However, such machines are expensive and heavy—they may weigh ten times as much as the person they are lifting—and will, therefore, probably be deployed primarily in nursing homes or hospitals. Building a low-cost robot with sufficient dexterity to assist with personal hygiene or using the bathroom remains an extraordinary challenge. Experimental machines capable of specific tasks have appeared. For example, researchers at Georgia Tech have built a robot with a soft touch that can give patients a gentle bed bath, but the realization of an affordable, multitasking elder-care robot that can autonomously assist people who are almost completely dependent on others probably remains far in the future.

One of the ramifications of that daunting technical hurdle is that, despite the theoretically huge market opportunity, there are relatively few start-up companies focused on designing elder-care robots and little venture capital flowing into the field. The best hope almost certainly comes from Japan, which is on the brink of a national crisis and which, unlike the United States, has little aversion to direct collaboration between industry and government. In 2013, the Japanese government initiated a program in which it will pay two-thirds of the costs associated with developing inexpensive, single-task robotic devices that can assist the elderly or their caretakers.
²⁰

Perhaps the most remarkable elder-care innovation developed in Japan so far is the Hybrid Assistive Limb (HAL)—a powered exoskeleton suit straight out of science fiction. Developed by Professor Yoshiyuki Sankai of the University of Tsukuba, the HAL suit is the result of twenty years of research and development. Sensors in the suit are able to detect and interpret signals from the brain. When the person wearing the battery-powered suit thinks about standing up or walking, powerful motors instantly spring into action, providing mechanical assistance. A version is also available for the upper body and could assist caretakers in lifting the elderly.
Wheelchair-bound seniors have been able to stand up and walk with the help of HAL. Sankai’s company, Cyberdyne, has also designed a more robust version of the exoskeleton for use by workers cleaning up the Fukushima Daiichi nuclear plant in the wake of the 2011 disaster. The company says the suit will almost completely offset the burden of over 130 pounds of tungsten radiation shielding worn by workers.
*
HAL is the first elder-care robotic device to be certified by Japan’s Ministry of Economy, Trade, and Industry. The suits lease for just under $2,000 per year and are already in use at over three hundred Japanese hospitals and nursing homes.
²¹

Other near-term developments will probably include robotic walkers to assist in mobility and inexpensive robots capable of bringing medicine, providing a glass of water, or retrieving commonly misplaced items like eyeglasses. (This would likely be done by attaching RFID tags to the items.) Robots that can help track and monitor people with dementia are also appearing. Telepresence robots that allow doctors or caretakers to interact with patients remotely are already in use in some hospitals and care facilities. Devices of this type are relatively easy to develop because they skirt around the challenge of dexterity. The near-term nursing-care robotics story is primarily going to be about machines that assist, monitor, or enable communication. Affordable robots that can independently perform genuinely useful tasks will be slower to arrive.

Given that truly capable and autonomous elder-care robots are unlikely to emerge in the near future, it might seem reasonable to expect that the looming shortage of nursing home workers and home health aids will, to a significant extent, offset any technology-driven
job losses that occur in other sectors of the economy. Maybe employment will simply migrate to the health and elder care sector. The US Bureau of Labor Statistics (BLS) projects that by 2022, there will be 580,000 new jobs for personal-care aids and 527,000 for registered nurses (those are the two fastest-growing occupations in the United States), as well as 424,000 home heath aids and 312,000 nursing aids.
²²
That adds up to about 1.8 million jobs.

This sounds like a big number. But now consider that the Economic Policy Institute estimates that, as of January 2014, the United States was still short 7.9 million jobs as a result of the Great Recession. That includes 1.3 million jobs that were lost during the downturn and hadn’t yet been recovered as well as another 6.6 million jobs that were never created.
²³
In other words, if those 1.8 million jobs all appeared today, they would fill only about a quarter of the hole.

Another factor, of course, is that these jobs are low-paying and not particularly suitable for a large fraction of the population. According to the BLS, home health aids and personal aids both provided a medium 2012 income of under $21,000 and require an education level of “less than high school.” Large numbers of workers are likely to lack the temperament necessary to thrive in these jobs. If a worker hates his job stamping out widgets, that’s one thing. If he despises his job caring for a dependent older person, that’s a major problem.

Assuming the BLS’s projections are correct and these jobs do materialize in large numbers, there is also the question of who will actually pay for these workers. Decades of stagnant wages, together with the transition from defined benefit pensions to often under-funded 401k plans, will leave a large fraction of Americans in relatively insecure retirement situations. By the time the majority of older people reach the point where they need personal, daily assistance, relatively few are likely to have the private means to hire home health aids, even if the wages for these jobs continue to be very low. As a result, these will probably be quasi-government jobs funded by programs like Medicare or Medicaid and will therefore be viewed as more of a problem than a solution.

Unleashing the Power of Data

As we saw in
Chapter 4
, the big data revolution offers the promise of new management insights and significantly improved efficiency. In fact, the increasing importance of all this data may be a powerful argument for consolidation in the health insurance sector, or alternatively creating some mechanism for sharing data among insurance companies, hospitals, and other providers. Access to more data could well mean more innovation. Just as Target, Inc., was able to predict pregnancy based on customer purchasing patterns, hospitals or insurance companies with access to large datasets will potentially discover correlations between specific factors that can be controlled and the likelihood of a positive patient outcome. The original AT&T was famous for sponsoring Bell Labs, where many of the twentieth century’s most important advances in information technology took place. Perhaps one or more health insurance companies with sufficient scale could play a somewhat similar role—except that the innovations would come not from tinkering in a lab but from continuously analyzing reams of detailed patient and hospital operational data.