The amount of globally produced data is at an all-time-high and continuously increasing. New applications and technologies are emerging with rapid speed and many of them rely on data for improvement. One field that has particularly been fueled by the increasing availability of data is computer vision. In combination with deep learning and artificial neural networks, applications that use computer vision have been adopted across many industries.
Mobility has been significantly enhanced by computer vision
One of the areas in which computer vision has been adopted rapidly is the mobility sector. Earlier vision systems were utilized for automated vehicle assembling. Ongoing technological advancements led to a wider scope in the industry to entail automotive driver systems and traffic management systems.
Computer vision systems that enable autonomous driving and other mobility services and systems require a large amount of video data and are therefore accompanied by numerous privacy-related debates and court rulings, aiming to ensure a lawful handling of personal data.
The algorithms which are used to detect the visual context of a vehicle’s surroundings are capable of more than automatically navigating through it. Within the last few years, the use of third-party dashcams in cars, motorcycles and bikes has increased. Visual data recorded during accidents and other prohibited traffic behavior can be admissible as court evidence.
Data privacy and computer vision must go hand-in-hand
The EU Council of Ministers passed a general safety regulation that requires OEMs to build new cars with advanced safety systems. The mandatory safety systems include “warning of driver drowsiness and distraction, intelligent speed assistance, reversing safety with camera or sensors, and data recorder in case of an accident”. Cars and vans must also be equipped with lane-keeping assistance and advanced emergency braking. The commission expects that by 2038, 140,000 serious injuries will be avoided and 25,000 lives will be saved.
To create such systems, it is inevitable to capture and process a wide variety of traffic situations as a representation of the reality in which vehicles operate. The indispensability of real data has been recognized and appreciated by lawmakers. The German Association of the Automotive Industry (VDA) and Germany’s federal and state data protection authorities have drafted a joint declaration to regulate the use of video data, derived from rides that are intended to collect data for scientific purposes and help to develop autonomous vehicles and increased safety systems.
The legitimate interest of companies to enhance driver assistance systems and autonomous driving serves as the legal basis for collecting and processing traffic video data. These developments are furthermore in public interest as they contribute to the increase of traffic safety. Of course, there are more regulations to be considered and taken care of, when collecting traffic data for the development of such systems.
The important message here is that regulators understand and appreciate the fact that the use of video data is a crucial factor for the future of mobility and transportation.
The collection of visual data is a great chance and a great challenge at the same time. Globally, private and public organizations have realized the importance of collecting visual data and the importance of collecting it in a privacy friendly way. Visual data is the fuel that is needed for technical solutions that are desirable for companies and governments alike: enhancing profitability, increasing safety and ecology, while ensuring public privacy protection.
At Peregrine we are committed to Vision Zero, while maintaining the highest standards of privacy by design.