IoT Car: The Internet of Things is Already Here and its called Uber

Smart Lifestyle

IoT Car: The Internet of Things is Already Here and its called Uber

The company that has revolutionized transportation also represents a template for building IoT car businesses.

by Alan R. Earls

Leveraging the rapid uptake of smartphones globally, the company created a marketplace for drivers and passengers that has made transportation faster and more affordable for millions while thoroughly disrupting traditional players, such as taxi companies. Hailing cabs the traditional way is a thing of the past for many. However, Uber has also pioneered what might well be seen as a giant Internet of Things enterprise; one that offers lessons and inspiration for other entrepreneurs.

Uber and airbnb are interesting because they are really internet of things platforms.

Tim O’Reilly
Technology guru and founder of O’Reilly Media

To be sure, most casual observers would be unlikely to classify Uber as an IoT company at all. As blogger, Dr. Jim Walsh, CTO at GlobalLogic, a product development services provider, notes in a post on Uber, we tend to identify the company with what it provides us – a comfortable and affordable way of getting places if we are a passenger, and a easy-entry business opportunity for drivers.

However, Walsh explains that once the Uber app is launched, “the location sensors associated with both the passenger’s and the driver’s mobile devices (the actual ‘things’ being monitored) are regularly broadcasting their location to a ‘back end’ system that is hosted by Uber in the cloud. Uber’s cloud service then provides analytics to determine which car/driver combinations are relevant to each service request. That’s not unlike the architecture of any other IoT system.
Quoted last year in the Guardian, technology guru Tim O’Reilly, founder of O’Reilly Media, put it more directly: “Uber and Airbnb are interesting because they are really Internet of Things platforms. They are able to catalyse a swarming marketplace management model because everyone is carrying around a connected sensor package.”

For Uber, “life” following this approach has been good, with the company expanding rapidly across the U.S. and globally in just a few years. Building on its success, the company has also tested out Uber-based online food ordering service, a package delivery system, and a car pooling service. Its successes have created like-minded companies, most of which operate similar services, in a process sometimes dubbed “uberfication.”
Given its mastery of this new kind of infrastructure, Uber may be contemplating a further strengthening of its links to the devices of its members in hopes of leveraging them for additional use – potentially tapping them as sensors for new kinds of business. “If you look at the Uber model, it is easy to imagine a company offering what one day in the future could be part of a smart city strategy – with Uber hypothetically going around and collecting data that could then be applied in ways only limited by imagination,” says John Jackson, program vice president for mobility research at IDC, an analyst firm.
And, he notes, there’s every reason to think future IoT opportunities could also be similarly loosely coupled; linking information from or about physical assets, locations, or activities to an intermediary service that enables transformative and profitable activities. Again, the sky is the limit.
Just think what IoT sensor data from “oberfied” agricultural activities in California’s Central Valley might be able to tell a railroad or trucking company about when to plan for peak harvestshipments. Vehicles could be deployed “just in time” from other geographies to ensure timely and efficient movement of perishable products. Similarly, IoT-enabled roadways could power services for autos, the supply chain, entertainment, or almost anything else that can be imagined given how central motor transport is to the economy of the developed world.

Sensor technology – a potent enabler

As both Walsh and O’Reilly have pointed out, the Uber model relies on the rich set of hardware and applications available on a typical smart phone. GPS is the most obvious, but many other sensors are already builtin and more will likely appear in the future. Thus, researchers at the Indian Institute of Technology, have proposed a traffic and road condition estimation system using smartphone sensors; a system they call Wolverine1.

They believe that better monitoring could improve traffic efficiency. Up until now, solutions that have been put forward have generally required investment in dedicated hardware such as GPS devices and accelerometers in vehicles aswell as roadside cameras. Instead, they propose simply harnessing the sensors present in smartphones – either directly measuring aspects of traffic flow or inferring events such as acceleration or braking. The net result would be an inexpensive, nearreal-time monitoring system that could potentially even help re-route drivers and vehicles in response to traffic jams.

Cautionary Voices – IoT Car

All of these initiatives and technologies are moving fast; too fast in the eyes of some. One warning comes from Lev Lesokhin, executive vice president of strategy and analytic at New York City-based CAST, a company focused on software risk prevention and analysis. “Broadly, there are two types of software; one is focused on information technology, the software that runs personal computers and businesses and connects the internet. The other kind of software is embedded software, which runs devices, machines, and sensors,” he notes. Those two families of software are very nearly separate industries because they have different standards and expectations. People working to program devices in automobiles have their own standard for quality, as do others in other parts of the embedded world. “In fact, in general the way embedded peopleapproach quality has tended to be more rigorous than with information tech,” he says. The motivation is obvious. If a personal computer has a glitch, or even if an Uber driver misses a message, the consequences are usually annoying rather than calamitous. Embedded devices, on the other hand, may control medical equipment or machinery or ensure the safety of vehicles, the failures of which could have life-threatening consequences. Thus, he noted, embedded products take more of an engineering approach to software design, development, and testing.

All of these initiatives and technologies are moving fast; too fast in the eyes of some. One warning comes from Lev Lesokhin, executive vice president of strategy and analytic at New York City-based CAST, a company focused on software risk prevention and analysis. “Broadly, there are two types of software; one is focused on information technology, the software that runs personal computers and businesses and connects the internet. The other kind of software is embedded software, which runs devices, machines, and sensors,” he notes.

Uber: Internet of Things - IoT Car

Photo: Uber

Those two families of software are very nearly separate industries because they have different standards and expectations. People working to program devices in automobiles have their own standard for quality, as do others in other parts of the embedded world. “In fact, in general the way embedded peopleapproach quality has tended to be more rigorous than with information tech,” he says. The motivation is obvious. If a personal computer has a glitch, or even if an Uber driver misses a message, the consequences are usually annoying rather than calamitous. Embedded devices, on the other hand, may control medical equipment or machinery or ensure the safety of vehicles, the failures of which could have life-threatening consequences. Thus, he noted, embedded products take more of an engineering approach to software design, development, and testing.

The consequence for those aiming to build a new world of IoT around embedded devices and sensors, though, is the necessity to fuse those two worlds together, making sure that both are “improved” in the ways needed to make a robust system.
“IoT changes the challenges for the embedded people, especially from an engineering and quality standpoint,” says Lesokhin. That’s because, when you are writing embedded software, by definition it has almost always been isolated at the device. So you have to work with the sensors and hardware on the device, and manage that – nothing more. On the other hand, IT-oriented software typically considers transactions and interactions between different pieces of the overall information systems. “If you think about any kind of cloud-based software or any system that runs organizations, you collect info in one place and send it to another to process,” he says.

Uber is a perfect example of information driven software.

Lev Lesokhin
Executive vice president of strategy and analytic at CAST

Another important concern for all of IoT is security. Indeed, Uber itself has suffered at least one substantial data breach in its brief history. Again, Lesokhin points out that security has always been the last thing on the mind of embedded developers. There is a huge need to update thinking and capabilities as IoT engages more embedded devices and systems. Last year Uber launched a private, beta bug bounty program for over 200 security researchers who ended up finding a total of some 100 bugs – all of which have since been fixed. As a follow up, Uber recently announced an official public “bug bounty” program. Payouts will go up to $10,000 for critical issues. Future Ubers will also need to tackle their own security challenges head on.
Performance is also an issue, he says, specifically network latency. When the network is a little slow are you managing your timeouts correctly? These aren’t things you think about on an embedded device.

“I think that is the fundamental starting point for what makes for reliability, security and performance of software on embedded devices,” he says.

Making it all Work

“I think that is the fundamental starting point for what makes for reliability, security and performance of software on embedded devices,” he says.
making it all Work Lesokhin says the cures for these system challenges are in the works or on the horizon. For instance, the Object Management Group (OMG) and the Software Engineering Institute at Carnegie-Mellon University have worked to put together a set of standards for creating software that is reliable. “I think that a set of standards like that is going to be an important part of making sure IoT is safe, sound and secure,” he says. “Hypothetically we are all in effect nodes for Google, Apple, Facebook, Amazon, and Uber and thus can end up through third parties in the IoT space,” says IDC’s Jackson. Today, Uber might not count as IoT under a traditional definition, “but these lines will blur,” he adds.

Alan R. Earls is a Boston-based freelance writer focused on business and technology.

1 Technical Paper Wolverine: ravi.bhoraskar.com/papers/wolverine.pdf