It's a given that cars will drive themselves someday – and some already can for the most part. It's also a given that, in addition to sophisticated cameras and sensors, one of the key pieces to the autonomous car puzzle is having highly accurate mapping software.
We already have that, too, with digital mapping detailing our world like never before. And as precise and more granular mapping software is combined with real-time traffic and other situational data harvested from connected cars, it will allow self-driving vehicles to not only stay on track, but also provide critical information on what's ahead – which can make traffic flow and the entire in-car experience much better. And perhaps even personal.
That was the takeaway after visiting the HQ of Nokia Here in downtown Chicago last week to get a first-hand look at the company's traffic centre as well as its R&D operations. Here, a division of Nokia, wanted to show how its mapping software and research into connected car data could someday be combined for what the company calls "Highly Automated Driving."
"Before drivers let go of the wheel, they must feel confident that their cars will keep them safe, and become comfortable with the way that the system drives," said Ogi Redzic, Here's vice president of Connected Driving. And for those worried about driving becoming as uniform and uninvolved as riding the bus or taking the subway in a future of robo-cars, here's the good news: The company's autonomous vehicle vision incorporates the type of individualisation that human drivers are used to – sans the traffic and accidents cars currently cause, on everything from a winding country road to China's busiest urban highway.
"Taking into account detailed road geometry, information like weather and road conditions, and sensor data, cars can predict how to drive based on how a person normally drives," Redzic added. "So, for example, the car will know at what speed to take a curve based on an individual's comfort level, how other drivers are doing it that day, and what drivers have historically done based on the weather that day."
Gathering all the data
In addition to developing high-definition mapping software that reveals challenging road features ranging from circular motorway off-ramps to right-angle hairpins turns, Here showed how it also gathers and processes vast amounts of real-time traffic data on how weather can affect drivers. In a large room at the top of a Chicago skyscraper, a small platoon of Here specialists each scan a half dozen or so computer monitors at individual workstations.
Each one handles a major metropolitan area and monitors online traffic reports, roadway webcams, and even transportation Twitter feeds. "We've found that the information that public agencies post on Twitter is usually more timely than what they provide through their other official channels," Mike Dekrell, senior manager of Here Traffic Operations, pointed out.
This is the type of real-time, crowd-sourced information that Here plans to leverage in the future to balance changing traffic conditions with driver behaviour "to create a more human autonomous driving experience," Redzic said. As an example, Jane Macfarlane, head of Here Research, showed a graphic (see below) that depicted the use of windshield wipers by taxi drivers in Eindhoven, Germany, after their cabs were outfitted with GPS and sensors.
Macfarlane explained how, when a thunderstorm passed through the city, real-time information on windshield wiper usage by the cabbies could be correlated with data on the average historical speed of vehicles in a city during similarly rainy weather. Self-driving, connected cars in the area can then incorporate this data to reduce their speed and keep traffic flowing more freely. (A recent Intel study showed most drivers are willing to have their cars controlled for the greater good of everyone on the road).
"Autonomous cars will not be one size fits all," Redzic said. "Cars will need to match the way that people drive in reality. That's where highly precise maps and sensor data combined with connectivity play roles in creating a more personal, human automated driving experience – one where the car becomes a co-driver that adapts to individual driving styles and preferences. Analysing how drivers behave with probe and sensor data and a map," he added, "Here learns how people prefer to drive – and ultimately how they want to be driven."