A new study of potential traffic patterns of driverless vehicles in New York is suggesting that the city would grind to a halt if 10% of vehicles are hacked and forced to stop moving.
Extremetech.com reports that Researchers from the Georgia Institute of Technology developed a model to predict what would happen as self-driving cars were attacked. They found you’d only have to take out a fraction of vehicles in a city to bring traffic to a complete halt.
It says the study used Manhattan as the basis of the model, and according to study co-author Skanda Vivek, a single hacked vehicle slows down the traffic in its immediate vicinity. As more vehicles go offline, these slowed areas merge and block the flow of traffic for cars that are still functional. Eventually, it says, you get gridlock.
It adds the team developed this model using a statistical approach called percolation theory, a mathematical tool that describes the behaviour of connected clusters.
The report says that if someone managed to hack just 10 % of rush hour vehicles, half of Manhattan’s 8,000 streets become inaccessible and by the time you reach 20%, the island is at a standstill. It adds that this is realistic, because if a carmaker like Toyota or Ford had an unpatched flaw that led to its autonomous cars being disrupted, that could easily be more than 10% of total vehicles on the road.