AfriForum, on Friday 23 April 2021, expressed concern about the South African National Roads Agency’s (Sanral) possible use of machine learning and facial recognition technology in the agency’s highway cameras.
The organisation now questions Sanral’s ability to implement the project successfully, citing the possible violation of the public’s constitutional right to privacy. Here’s what we know.
Sanral’s Technical Innovation Hub (TIH) earlier this week confirmed it was exploring the use of machine learning to improve road safety, reduce traffic congestion and assist with infrastructure development.
At the time, TIH mechatronic engineer Ruan van Breda explained the various benefits of machine learning, such as being used to “detect and segment and classify objects within a camera frame based on pre-trained image classifiers”. He adds:
“Within the road environment, this allows one to detect and classify different types of vehicles, pedestrians, animals or cyclists, among others. There is already ample data available to classify objects according to these types”.
Van Breda explains that the system adapts and learns as more data is collected. He admitted that this type of tech was still in an “exploratory phase in South Africa” and cited countries such as China using machine learning to assist law enforcement.
“[Authorities in China] are able to identify the driver of a vehicle and instantly issue fines if that driver does not have a valid driver’s licence. Fines can also be issued automatically for individuals who jay-walk or gain access to restricted areas”.
If implemented properly in South Africa, this technology could aid Sanral with infrastructure planning of drop-off or pick-up points, and identify which areas are considered a safety risk.
AfriForum believes this technology would violate the public’s constitutional right to privacy. The organisation called on Sanral to provide detailed information on how the tech would be implemented in South Africa.
AfriForum also requests information on what steps Sanral would take to ensure motorists’ privacy are protected, and how Sanral would combat misuse of this technology.
Legal and Risk Manager Marnus Kamfer explained that AfriForum “is not opposed to the use of the latest technology to improve road safety and law enforcement”, as long as it isn’t implemented “at the expense of people’s constitutional rights”.
“AfriForum is also concerned that Sanral refers to China as an example where [machine learning] has been successfully implemented, especially due to the Chinese government’s poor track record in terms of human rights and privacy violations”.
AfriForum says technological advancements cannot be pursued by a government to increase its ability to observe “and thereby infringe on the privacy of its citizens, especially not in a way that can lead to unhindered abuse”.
Machine learning is a branch of artificial intelligence (AI) that focuses on applications (or algorithms) that learn from any given experience, finds the pattern in massive amounts of data, and then improves its decision-making accuracy.
In the simplest terms, machine learning is the use of algorithms to identify and apply patterns in digitally stored data. This may sound dull, but machine learning, in essence, runs the world.
You know when Facebook shows you sponsored content and ads, or when Netflix recommends a series or film? That’s machine learning. Therefore, machine learning is used in everything from social media and Google search engines to your iPhone’s Siri or the ‘autoplay’ feature on YouTube.
The platforms “learn” your behaviours by collecting as much data about you as humanly (machinely?) possible. It then spots the patterns in your behaviour from the data collected and applies that pattern.