The business of transportation is as old and new as civilization, from moving crops to market to seafaring bulk carriers to global air travel. It is a trade that also includes the mundane morning commute.
All that moving of people and goods takes labor, including more than 3.5 million professional truck drivers in the United States, as well as 665,000 bus drivers, and 119,000 airline and commercial pilots. From the early ox-carts to jumbo jets, only a human could reliably guide transport.
But within the next decade the human may be supplanted behind the wheel by cheap cameras and sensors, artificial intelligence and GPS. The day of the autonomous vehicle, vessel or airplane appears nigh.
Vast fortunes will be made by enterprises properly positioned to benefit from the migration to driverless technologies. As in the early days of the dot-com era, one might suspect some money will also be lost.
Too, skeptics abound, pointing to midcentury or even later before the driverless delivery drone or taxi-van pulls up in front of your house for a routine pick-up or drop-off. For example, the British Columbia–based Victoria Transportation Policy Institute in January estimated that through the 2020s autonomous-option vehicles will represent only 2 percent to 5 percent of North American motor sales.
Possibly so, yet sober corporate players are venturing large sums to bring the autonomous world closer. For example, in mid-February motor giant Ford reported it would spend $1 billion to finance Argo, a Pittsburgh-based startup, to develop driverless technologies. Ford said it is targeting a 2021 commercial introduction of autonomous vehicle fleets.
And Ford is hardly alone venturing off-the-beaten-road, as both Toyota and GM previously each allocated $1 billion in driverless technology ventures.
“The race to build the fully connected car, and ultimately the completely autonomous vehicle, is already under way,” reported Big Four accounting and consulting giant PWC in a late 2016 industry study.
For investors, PWC’s observation is both encouraging and yet cautionary. Races have winners — but also losers.
The number of publicly held companies that are primarily driverless technology plays is small, and maybe even just one: Mobileye, an Israeli startup that raised $890 million in a 2014 IPO. Mobileye designs cameras, sensors and software to help prevent or mitigate vehicular accidents, as in braking automatically and rapidly before a human driver can.
The stock is pricey at 114 times trailing earnings, but the company is already profitable in a sustainable business model, trades at a more reasonable 44 times future earnings, and revenues are growing. In February, Mobileye reported a partnership with the world’s largest automaker, Volkswagen, to create a navigation system for autonomous vehicles. If autonomous vehicles ever become common, almost certainly they will rely on cameras and other sensors, constantly updated maps, cloud connectivity, and excellent onboard software, all of which play to Mobileye strengths or alliances.
Mobileye also appears well positioned to profit from ongoing safety trends in the motor-vehicle industry. For investors there is the comfort that Mobileye might make money on what could be the long road to self-driving cars.
NVIDIA is better known for designing graphic-processing units for the video gaming industry, but happily enough the technology lends itself to image recognition in real life as well. When Swedish automaker Volvo put 100 self-driving cars on the roads around its hometown of Gothenburg, the company selected the NVIDIA Drive PX 2, a supercomputer for vehicles that uses deep learning, sensor fusion and surround vision. Like other autonomous-vehicle proponents, NVIDIA notes that 94 percent of vehicular accidents are due to driver error and posits that fleets of self-driving vehicles would actually be safer than the human-controlled variety. Due to long immersion in the relentlessly competitive video-gaming industry, NVIDIA may have an upper hand in developing the artificial intelligence and image recognition vital to vehicle autonomy.
After 15 years of not doing much for investors, NVIDIA stock was a high-flyer through 2016, but it has been cooling off in 2017. With a market cap of more than $54 billion, and a 52-week trading range between $31 and $121, NVIDIA explains why buying autonomous-vehicle stocks is for the intrepid, and, shall we say, self-directed investor.
The Britain-based Delphi supplies to the auto- and truck-industries “OEMs” or original equipment manufacturer parts, so it has a larger business model than just autonomous vehicle parts and technology. Yet the Singapore Land Transport Authority chose Delphi to lead a consortium to put six self-driving electric-power taxis on clogged city-state streets to carry riders to and from mass-transit stations. It is a challenging real-world test to be carried out in 2017–2018, and if Delphi succeeds, it likely would be positioned to offer similar service to other cities. Delphi officials have issued confident statements, but so far the system is not yet tested.
The Delphi-Singapore outlook introduces an interesting angle to the autonomous vehicle story: The successful companies may be those who win and deliver on contracts, even if their technology is not best-in-trade. Graybeards will remember when IBM became Corporate America’s default computer in the 1960s and 1970s, despite other companies having perhaps better equipment. IBM was known for backing up its products and its solid marketing efforts.
THE UBER MODEL
The ride-sharing service Uber has hinted at a 2017 IPO, and if the taxi outfit can commercialize self-driving robo-cars, its upside and that of other similar taxi companies could rival successful web enterprises.
“The rise of fleets of shared robo-taxis will lead to a sharp drop in passenger price per kilometer. Robot cabs with battery-electric propulsion will be able to drive themselves to inductive charging stations ...,” wrote PWC in its industry report.
By most accounts, battery vehicles are low maintenance and cheaper to operate on a per mile basis than fossil-fuel cars. The upfront battery costs — the last hurdle to affordability — are being continuously lowered as well.
Reasonable projections are that Uber vehicles could travel more than 300 miles per day on an off-hours charge, which is plenty of miles enough for urban commuting. In another advantage, a regional Uber taxi vehicle would need only negotiate local terrain and hazards, and need not have accurate global maps at the ready.
There are private, pre-IPO shares of Uber rarely on the market, perhaps available through such website-markets as EquityZen or Sharepost, though Uber is said to be policing the availability of shares strictly. For Uber wanna-be investors, finding a way to buy Uber at the IPO price (probably by being a good customer at an underwriting brokerage) is the next-best tack.
In late February, Uber was sued by Google’s parent Alphabet, alleging their key sensor technology was lifted by a former employee from Waymo, an Alphabet company.
The suit highlights a strength and a weakness of Uber: Like others in the space, it may or may not have the best technology to develop a self-driving vehicle — evidently a crowded field. But perhaps for Uber, owning technology is not so important as having a fleet of self-driving battery cars. Uber can make money by having the best-known brand in the lowest-cost taxi-service category, a need in every city globally.
If hailing a robo-taxi, even daily, becomes cheaper than owning a car, and customers can relax, read a tablet or talk on a smart-phone while being automatically chauffeured across town, then the Ubers of the world likely have a future with legs … or, should we say, wheels.
ON THE FARM
The farm in some regards is tailor-made for autonomous vehicles. Agricultural plots are usually private property with static boundaries and hazards. With the use of GPS and modern sensing technologies, farm equipment is already programmed to semi-autonomously carry out routine tasks, such as plowing and seeding in precise rows, with the tractor driver negotiating only the end-row turns.
Case IH, major European-based agriculture equipment and vehicle maker, prototyped a completely driverless tractor during 2016 and already installs technologies into its tractors that allow semi-autonomous tractor driving.
Unfortunately, with large mechanized farm equipment, tragic accidents happen even with drivers, sometimes involving children in fields. A Purdue University study noted that a single tragedy could easily turn a farm community against a large autonomous vehicle, and that farms might have to be strictly fenced, re-creating the secure perimeter of factories, which today, in fact, routinely deploy robot labor.
Through history, agricultural implements have adapted to farms, and vice-versa. The competitive farm of the future may be designed and shaped to take advantage of autonomous vehicles, which are also able to test and amend soils as they plow and plant. The term “industrial farm” might yet give way to “robo-farms.”
Militaries have a reason above and beyond dollars and cents for deploying autonomous vehicles: force protection, or saving troop lives. Resupplying frontlines can be hazardous work, often requiring exposure to enemy attack. Oshkosh manufactures the autonomous or semi-autonomous TerraMax, a vehicle designed as a military transport. U.S. Army officials say they are looking to deploy autonomous vehicles in convoys within 10 years.
Blueprints of the Oshkosh vehicle reveals how complicated autonomous vehicles can be, as the Terramax transport not only relies on multiple radars, it also uses infrared cameras, LIDAR (light detection and ranging sensors, utilizing laser beams) and optical cameras, in addition to GPS and also military-satellite location systems. All of that, in one truck.
Autonomous vehicles will also have a place in military reconnaissance. In 2016, the U.S. Navy and DARPA revealed they were operating a semi-autonomous drone sub-hunter. The idea behind the vessel Sea Hunter is that it relentlessly sails about selected waterways, with its sonar and other sensors pointed down and searching endlessly for submarines — the kind of monotony best left to machines. The vessel will automatically react and take action once a sub is located.
Defense engineering contractor Leidos designed key elements of theSea Hunter. The $7 billion-in-sales company also builds an autonomous underwater vehicle for the U.S. Navy, which looks like a torpedo and can collect information. Underwater vessels may need autonomy, as signals can be difficult to send or can be intercepted, thus revealing position. There can be tactical battlefield needs for “radio silence,” that increase the desirability of autonomous submarines.
Leidos, with 33,000 employees, is much larger than a single contract, or even two programs for the U.S. Navy. Yet Leidos has emerged as a leader in a field that appears poised for inevitable growth, the robotization of military or security operations. People are expensive, become bored, and the risking of a machine pales next to the risking of a human life.
OUT TO SEA
Container and tanker ships criss-cross the globe but — even today — at the speed of a horse-drawn carriage, all the while eating up crew hours for every league logged. By some measures, labor accounts for 44 percent of cargo-shipping bills.
The company Rolls-Royce may conjure up images of luxury cars or large jet engines, but the storied manufacturer is the lead corporate partner in the Advanced Autonomous Waterborne Applications Initiative, funded by Tekes, which is Finland’s funding agency for technology. The project brings together maritime businesses and Finnish universities, and has predicted autonomous cargo ships might be a reality within 20 years.
The first step will be semi-autonomous but also remote-controlled ships, perhaps in a little more than a decade, predicts Rolls-Royce. As with driverless cars, the major tangles ahead might be more legal and jurisdictional rather than technological, given the multiplicity of nations and authorities involved that govern shipping, especially close to shore. But if robo-ships ever ply the high seas, Rolls-Royce might be at the helm.
AUTONOMOUS JUMBO JETS
It is an impolite topic, but one of the causes of recent commercial airline disasters has been pilots, and not because of human error. In recent times, the crash of GermanWings Flight 9525 in 2015; the vanishing of Malaysian Airlines Flight 370 in 2014, which mysteriously remained aloft and functional for many hours after contact was cut off; and the crash of Egyptian Flight 990 in 1999 have been laid to intentional pilot actions.
Unbeknown to most passengers, large commercial airliners today are already usually guided from the ground or autonomously by computer autopilot, with the human pilot assuming command but for spare minutes during a typical flight.
Of course, the famed US Airways Flight 1549 landing in the Hudson River in 2009, and then the Clint Eastwood–directed movie Sully has reinforced in the public’s mind that there had better be a smart and competent human being in the cockpit for air emergencies.
Still, others see a need for ground control to interdict and automatically guide errant commercial aircraft, even overriding crew.
The issue of jumbo airliners also raises an interesting intellectual topic: When is a vehicle, plane or ship autonomous, and when is it a drone? The future will likely have drone vehicles that are usually autonomous, but then also guided occasionally, either by human or automated ground control.
Is an airplane guided by automated ground control a “drone,” or still classified as “autonomous”?
If technological past is prologue, the winners in the autonomous-vehicle race may not be the swiftest, but rather the best marketers or the most user-friendly. Google did not invent the Internet or even the search engine, but it did make among the best and easiest to use search services. Apple products are famous for ease-of-use.
Taxi, trucking and shipping companies may emerge in coming years offering robo-service, and for lower rates. Some enterprises may win by developing the technology to allow autonomous vehicles. Others may win by making self-driving services usable for businesses and consumers.
Benjamin Cole is a freelance writer based near Korat, Thailand.