How should autonomous vehicles handle privacy?

By Catherine
Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+

By Catherine Bolgar

The development of autonomous vehicles raises a host of questions about data collection. Some of the issues may arrive as vehicles incorporate more automated systems and become connected to each other and to infrastructure—before they are fully autonomous. Here are some questions likely to arise.


What kind of data might be gathered about vehicles’ movements? By whom?

Fleet operators would likely have some data on their customers, just as the current ride-sharing services already do, says Kara M. Kockelman, professor of engineering at the University of Texas-Austin. “If I order a car, I will have to have a credit card on file for trip charges,” she says. “People already are giving that information up.”

Worries about unauthorized credit-card data replication would be the same as for current car services.

Insurers may want to collect or have access to data to establish liability in case of a crash, to determine weather or road conditions or whether the technology malfunctioned. California law requires autonomous vehicle makers to have a way to capture and store sensor data for at least 30 seconds before a collision and to keep it for at least three years after an incident.

California also requires that autonomous car makers disclose to buyers what information is collected by the technology on the vehicle, although it doesn’t specify whether the data would be gathered by the manufacturers, government agencies or private companies, or a combination.

For example, “location data, such as the day and time, and which road you’re driving on, would be valuable for transportation planning and understanding where the traffic is,” says Frank Douma, director of the state and local policy program at the University of Minnesota in Minneapolis.

A company collecting the data might want to sell it – already a common practice. “People are probably clicking through consent agreements far too quickly and don’t know if the company will share or sell the information,” Mr. Douma says.

People are putting their information on the open market without realizing it.”

Theoretically, home thieves could tap the data to monitor when occupants have left, but the experts say that’s probably more complicated than what burglars do now, which is just watch target houses in person.

Autonomous vehicles will use many kinds of sensors and technology, such as cameras, radar, LIDAR and GPS. Regulators may allow a vehicle’s GPS to keep track of itself but not share it with anybody, or to change identifiers, to preserve anonymity, Prof. Kockelman says.

The technology for connecting vehicles to each other, called connected car or dedicated short-range connection (DSRC), is under review by the U.S. National Highway Traffic Safety Administration (NHTSA). DSRC would use a bandwidth exclusively reserved for vehicle-to-vehicle communication.

DSRC technology “has built into it anonymization of information,” Prof. Kockelman says. “One vehicle will transmit to other vehicles its speed or whether the brakes go on, but it won’t identify itself as a particular vehicle owned by a particular person.”


Will autonomous vehicles be safe from hackers?

While autonomous vehicles are expected to be safer than human drivers, cybersecurity is a concern, as demonstrated by hackers who already have taken control of cars through existing connected systems.

“DSRC is important because it’s much more privacy-protected and secure than if the data were sent over the Internet,” says Dorothy Glancy, law professor at Santa Clara University School of Law in California. “It’s a dedicated network, a closed Internet of vehicles instead of the Internet of Things. That’s the debate right now: whether everything should be connected to everything else.”

Global Automakers, a worldwide industry group, created an Information Sharing and Analysis Center to assess cybersecurity in vehicle electronics. In addition, the European Automobile Manufacturers Association has agreed on secure principles of data protection for connected vehicles and services.


Who will update the maps?

One of the ways autonomous vehicles know where to go is through digital maps. It still isn’t known whether maps will be updated in real time—the way GPS applications currently readjust routes according to traffic conditions or construction work—or hourly or daily, or a combination.

A missed update could theoretically send an autonomous car into a construction zone, though some autonomous vehicles already are able to handle hand signals and flashing arrow signs. The bigger risk is likely to be bad weather conditions, such as snow, that cover lines on the road.

A consortium of European car makers acquired a digital map company that uses wireless transmissions to and from vehicles for updates. Other companies also have advanced mapping capabilities, protected by patents and trade secrets.

“Policymakers want these companies to get together and pool their information. In fact the very first item in the new ‘Federal Automated Vehicles Policy’ Vehicle Safety Assessment is ‘Data Recording and Sharing,’” Prof. Glancy says. “One of the features of the NHTSA guidance is for vehicle manufacturers to share vehicle performance data. Moreover, there’s no reason why you have to have two or three companies collecting the same information on the same roads. You want the most accurate and the most timely for all autonomous vehicles. Will that be a public function? Will it be done by municipalities or states? Will there be sharing across state lines? Who will hold the pool of mapping data?”

Like the autonomous vehicles themselves, answers to these questions remain largely a work in progress. Many stakeholders are working hard to enable the answers, especially on the technology that will connect autonomous cars with infrastructure and make mobility far safer.


Catherine Bolgar is a former managing editor of The Wall Street Journal Europe, now working as a freelance writer and editor with WSJ. Custom Studios in EMEA. For more from Catherine Bolgar, along with other industry experts, join the Future Realities discussion on LinkedIn.

Photos courtesy of iStock

Wearable Sensors Make Workplaces Safer

By Catherine
Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+


By Catherine Bolgar

close up of hands setting smart watch

Millions of people world-wide wear devices to track their activity, heart rate and sleep. The number of wearable sensors is expected to reach three billion by 2025. Employers are looking at how the devices can improve workplace safety, especially in dangerous or remote environments.

A hospital in Turin, Italy, experimented with a variety of wearable sensors as well as sensors incorporated into the environment—the Internet of Things, says Guido Boella, professor of computer science at the University of Turin, who was one of the researchers of the study, called Sensing Safety at Work. A dashboard showed  not only information from the sensors, but also a map of the hospital to pinpoint workers in distress.

Sensors on wristbands were given to employees in a laboratory where toxic substances were handled. “If they got sick or fell down, it set off an alarm and showed them on the map,” he says.

Another aspect of the study used sensors on bands on the wrist or body, or on necklaces, designed to avoid being invasive while employees worked, but which could signal problems by setting off an alarm via vibration, light or sound. A third experiment combined information about the dispersion of toxic substances in the air with the physiology of individuals.

Female scientist working in lab 7“We combined data from people with wearables with data from Internet of Things sensors”  detecting substances in the air, Dr. Boella says. “For many substances, the problem isn’t just a threshold level present, but how much of it is absorbed. That depends on a person’s activity, how long they are exposed, how high their heart rate is. The values from the body were combined with values from the Internet of Things sensors in the room to provide personalized alarms.”

Both the cost and size of sensors have diminished greatly in the past few years, but battery life is still limited, requiring either better batteries or more efficient hardware, says Saul Delabrida, assistant professor at the Federal University of Ouro Preto in Minas Gerais, Brazil.

Connections also are an issue. “The current state-of-the-art of the technology allows creating wearable devices with enough power-computing capabilities to process the data collected from the sensors without a Wi-Fi connection,” he says. “The sensors need to be connected to the module [that’s] able to execute an algorithm that provides information to the user in this strategy.”

Sensors can monitor workers for fatigue, overheating and cold, as well as location—to find a forestry worker alone in a remote place, for example. Already the mining industry uses sensors for detecting dangerous levels of dust and gases, and proximity sensors on both trucks and miners to reduce accidents. “When he raises a bucket of rocks, the driver doesn’t have a good view,” says Martin Lavallière, professor of health and kinesiology at the University of Quebec at Chicoutimi, Canada. An alarm in the truck warns the driver that a miner is nearby, while a wearable on the miner alerts him that he’s too near the truck.

Employers also can promote health and wellness through the activity trackers that many employees already wear, in a trend that reflects the BYOD—Bring Your Own Device—model for smart phones at work, Dr. Lavallière says, or they can choose one brand and provide it to all employees. Either way, the goal is to give incentives to employees, to make sure they get enough activity and maintain good health behaviors.

“If employees maintain their health they will be able to be more productive and more efficient, with fewer absences for sick leave, less ‘presenteeism’—when they show up for work sick and are not as productive,” he says.

Woman carpenter using circular saw.Employers could use the data on an individual level to encourage wellness, or could aggregate the information to gain insights into how employees react to situations at work. For example, “they could see that every time employees go to this workstation, their heart rate goes up,” Dr. Lavallière says. “Is that because the work is physically demanding, or because someone nearby is creating a stressful environment through their behavior, like yelling at people? When you look at the data from all the employees, you can say, ‘This job seems more stressful. How can we make it better? Should we separate the task in two?’”

Another question will be how employers deal with worker privacy and control.

The factory of the future could use these devices not to track whether someone is not feeling well, but whether they are making the right movements with their hands,” Dr. Boella says.

A wearable sensor could detect whether a worker is assembling an appliance in the authorized way. On the one hand, such monitoring could reduce repetitive-motion injuries. On the other, “there may be no more freedom to do things in a different way,” he says. “Not even controlling our bodies and movements.”

Tracking how many steps an employee takes or how long they sleep “could be seen as Big Brother at work,” Dr. Lavallière says.

In introducing wearable sensors in the workplace, employers need to explain the nature of the data collected and use incentives rather than punishment, he says, adding, “There will always be a conflict between the employee and employer. What’s important is that everyone draws benefits from the technology.”



Catherine Bolgar is a former managing editor of The Wall Street Journal Europe, now working as a freelance writer and editor with WSJ. Custom Studios in EMEA. For more from Catherine Bolgar, along with other industry experts, join the Future Realities discussion on LinkedIn.

Photos courtesy of iStock

Sensors in the Age of Industrial IoT

By Catherine
Share on LinkedInTweet about this on TwitterShare on FacebookShare on Google+

By Catherine Bolgar

The use of sensors, especially in the Internet of Things, is creating a mountain of big data. This infographic illustrates how to manage the data generated from sensors, including best practices for identifying which data streams are useful as the volume of data continues to explode.

Click here to see a larger view


Catherine Bolgar is a former managing editor of The Wall Street Journal Europe, now working as a freelance writer and editor with WSJ. Custom Studios in EMEA. For more from Catherine Bolgar, along with other industry experts, join the Future Realities discussion on LinkedIn.