The Internet of Things (IoT) has already begun the process of “digitizing the physical world,” in the McKinsey Global Institute’s phrasing. Right now, billions of interconnected sensors are busily spouting data on acceleration, temperature, location, power consumption and other variables, with minimal human intervention. IDC predicts that 30 billion devices will come online by 2020—close to four per person, using UN world population forecasts.
Soon enough, machines will generate more data through their own passive operations than humans will through active use. IDC estimated that in 2012, machine-generated data accounted for 30% of the “digital universe”; in 2011, the figure was 24%. The digital universe is itself ballooning, with a tenfold increase from 4.4 zettabytes in 2013 to 44 zettabytes expected by 2020. To say that this is equal to 44 trillion gigabytes is impressive, but abstract. Put another way, 42 zettabytes would be sufficient to store audio recordings of every word ever spoken by every human being who has ever lived.
Those who manage to organize and analyze this ocean of information will be able to gain ever more exquisite insights into the functioning of bodies, cars (a few of which already have all the sensor-generated information they need to run autonomously), homes, factories, supply chains—just about anything. This will in turn help them to reduce waste, prune inefficiencies, scope out shortcuts and generally improve the way systems operate. When it comes to the IoT, the axiom “knowledge is power” is more than a metaphor.
The Economy of Things
In a report for the IBM Institute for Business Value, Veena Pureswaran and Dr. Robin Lougee have taken a crack at characterizing and quantifying what they call the Economy of Things, the business end of the IoT. If the IoT is “digitizing the physical world,” then the Economy of Things represents the “liquification of the physical world.”
Pureswaran describes the thought behind that terminology: “the hypothesis here was that several industries have benefited greatly from digitization and the internet, and it’s very likely that the internet of things will bring that same benefit to the physical world, to industries that have typically had greater physical constraints, suffered from much more obstruction of information, and have typically lacked this idea of a more liquid marketplace.”
By way of historical analogy, the report detailed the effect of digitization on the airline industry. Partnering with American Airlines (AAL), IBM (IBM) began developing the Semi-Automated Booking & Reservations Engine (SABRE) in 1953. The effects of this innovation were muted prior to regulatory changes in 1978, but from that point on, air travel became “one of the first markets where every physical asset was digitized and placed into a single global online marketplace,” in the words of the report. Air fares roughly halved between 1978 and today. Planes that flew half-full for 6-7 hours a day now fill 85-90% of seats and fly 14 hours per day.
This transformation of the airline industry provides the conceptual framework for Pureswaran’s and Lougee’s study. For them, the Economy of Things will have three broad effects: “creating asset marketplaces,” “managing risk” and “improving efficiency.” SABRE highlights at least two of these, creating marketplaces for seats and improving efficiency by filling more seats and spending more time in the air.
The authors were not content to rely on analogy, however. “To understand this better we wanted to be able to model this quantitatively,” says Pureswaran, “so we worked with the Oxford Economics Institute on the actual economic modeling. The goal of the modeling exercise was to understand what the impact on different industries will be. We actually selected five industries for our case study. Three of them made it to the report.”
The three industries that appear in the paper are commercial real estate in the US, small and medium business lending in South Africa and agriculture in the US. The other two are SMB lending in Latin America and trucking and haulage in the US. Data assets are available to download for all five models (zip), including explanations of assumptions and adjustable parameters that give some color to the forecasts.
Creating Asset Marketplaces: US Commercial Real Estate
Only around 67% of the US’s 12 billion square feet of commercial real estate is utilized. The IoT, however, promises to make office space taggable, indexable and searchable in real time, so that an empty conference room can become a classroom, for example. The practice of hot-desking, in which there are fewer desks than employees—Deloitte’s new hyper-green Amsterdam office has only 4 desks for every 10 employees—could further improve efficiency through the process Oxford Economics’ Andrew Tessler calls “uberization” in the study’s data assets.
These and other adjustments could lead, the authors forecast, to a 39% increase in virtual capacity, based on a 50% adoption rate of IoT technologies. This in turn suggests a 42% drop in rents and a $128 billion net industry benefit. Lessors would not be so lucky, losing an estimated $14 billion.
Managing Risk: SMB Loans In South Africa
In South Africa, less than half of small and medium businesses (SMBs) operate in the formal economy. Since it is difficult to assess informal businesses’ credit risk, banks are reluctant to lend to them, and many informal SMBs therefore resort to higher-priced lenders or forgo borrowing altogether. The IoT could invigorate this sector, according to the report: “Combining device instrumentation, digital money, GPS logs and social networks, it will be possible for financial institutions to build much more accurate pictures of risk and simultaneously reduce the cost of repossession.”
Assuming 50% adoption of IoT technologies, the results could be dramatic: a $10 billion credit infusion could drive average interest rates down by 1% and boost GDP by 0.8% by 2020.
Improving Efficiency: US Agriculture
Farmers everywhere must contend with fickle odds. High-tech fertilizers and pesticides, genetically spruced-up seeds, fancy equipment and libraries worth of knowledge about soil and weather patterns help, but they did not stop US average corn-for-grain yield from swinging by 39% (upwards, thankfully) between 2012 and 2014.
Dr. Lougee, who contributed this section to the report, writes that by analyzing real-time data from sensors and drones, farmers can apply site-specific treatments, a process that could in large part be automated. Given the abundance of arable land in North America, US farmers have focused more on maximizing output per worker, rather than per acre. If the US were to catch up to northwest European yields through IoT innovation, however, food prices could drop 6% by 2020, boosting real GDP by 2%.
The question is, when all of this value is unlocked, who benefits? The answer, to the extent anyone knows, is not so simple. Take the historical precedent we began with, SABRE. Huge gains in efficiency did not translate into prosperity for airlines; collectively, the industry has not turned a profit since 1978. Most have filed for bankruptcy at one point or another. When American Airlines spun SABRE off in 1999, it was worth more than the parent company.
This would indicate that those who develop the data-processing tools that undergird the Economy of Things will reap the greatest reward. They certainly have their work cut out for them. McKinsey’s report on the IoT shows just how much data is typically lost before it can contribute to the decision-making process. Surveying an oil rig, they found that 40% is simply not stored; only about 1% is streamed onshore; this tiny subset is then carved up into a few metrics, which are analyzed retrospectively, not in real time.
Looking at this funnel, one marvels at the wasted potential of these 30,000 sensors. What if their garbled data could be effectively tagged, organized and analyzed in real time? IBM, AT&T (T), Cisco (CSCO), GE (GE) and Intel (INTC) would like to find out. In March 2014 they founded the Industrial Internet Consortium to explore real-world applications for the IoT, develop standards and best practices, address security gaps and facilitate the exchange of ideas related to the IoT.
There may also be an opportunity for chip makers, who are worse for wear due to flagging global demand. The SPDR S&P Semiconductor ETF (XSD) is down 7% year to date and nearly 19% this quarter. Sales may not recover up for a couple of years yet, but long-term prospects are brighter: PWC projects a 10.4% compound annual growth rate for sales of sensors and actuators through 2019; this is the segment of the semiconductor industry most associated with the IoT.
As with the regular-old internet, though, much of the windfall from the IoT is likely to be diffuse. As Pureswaran puts it, “At a macroeconomic level, we all end up winners.”
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of NASDAQ, Inc.