In a recent letter, A US traffic regulator has mentioned that the Artificial Intelligence controlling Google’s driver-less car can be considered a driver just like a human. Undoubtedly, this is going to give a big boost the self-driving car initiatives. As acknowledged by the director of Google’s self-driving car project “[it] will have major impact” on its development.
This year’s CES show also had some major announcements from companies like Ford, General Motors, Volkswagen, and Toyota in the space of connected cars, smart cars, and automated cars. No doubt these machines are going to be available on the roads pretty soon – within a matter of few years.
Where Does the Data Come from?
Such automated cars come with a countless number of sensors and gather a humongous amount of data. Google’s self-driving car gathers 750 megabytes of data per second! With embedded computers, GPS receivers, wireless networks, in-car sensors and the Internet, the future smart cars are going to be part of huge data-collection engine.
Sensors, actuators, and processors are the three main hardware components in the driverless car. Sensors gather the information and images and then the processor does the processing of that information to define the actions for the car via actuators. Actuators allow a computer to control the components of the car like brakes, or steering wheels.
Some other kind of data also comes from the surrounding environment and other cars. Cars with sensors, cameras, and radars detect objects around them and depending on the velocity of nearby objects, react on turning or sudden braking. Algorithms and predictive models help data-exchanging cars to foresee future events and proactively take actions based on that. GPS systems, cameras, short-range radios constantly exchange data with other cars to provide a safe driving experience.
Sensors also have the potential to identify unusual events and based on that, proactively take actions to avoid the potential performance issues. For example – the car can inform the car owner that a problem could possibly happen and it can also assist the owner in making an appointment with the nearest car maintenance shop.
Management and Processing of the Data
Considering the number of sensors and other data collection devices, it is quite believable that the smart cars are going to generate a gigantic amount of data. Overall, it could be several petabytes of data per year. Since every bit of information from every car may not be important, each byte of data is not going to be stored on the Cloud.
Several big auto giants have already set up big data centers to gather, store and analyze such data. Toyota, with its Data Communication Module (DCM) initiative, plans to create a Toyota Big Data Center to analyze and process data from its DCMs.
Ford has a lab in Silicon Valley which works towards using big data to improve its cars. Ford gathers data from over 4 million cars with in-car sensors and analyzes this data to improve fuel consumption, safety, quality and emissions in its cars. Engineers analyze this data in real-time to know and solve issues and improve their cars.
Concerns
Regulations and safety in the automated vehicles are some of the main concerns behind autonomous cars. Security risk and personal data-related problems could make such cars more hackable. Lack of strong privacy protection in the collection and processing of location data could result in regulatory issues against the technology and hamper the innovation. To avoid any undesired consequences, protection of location data should be built upfront into technologies, algorithms, and processes, while allowing the users to set up their privacy settings regarding their data.
Nonetheless, the future of automated cars is breathtaking. One of the most important benefits of automated driving is going to be improved road safety. With the use of right data and algorithms, robots might just be able to mimic the human mind – eventually even be able to take split-second decisions like a human.