While driving in Scotland on the A9 between Dunblane and Inverness during a recent vacation, I noticed a traffic sign that is unfamiliar to most North Americans: an average speed enforcement sign, pictured on the left. Instead of taking a picture of cars that exceed the speed limit at a particular point, an average speed enforcement system takes pictures of a car at two points kilometres apart, and calculates the average speed that the vehicle has driven between these two points. Average speed is distance divided by time, and this can be checked against the posted speed limit. Average speed systems use state-of-the-art automated number plate reading technology. As this technology is getting cheaper, expect such systems to appear in Canada and the United States as well. Average speed control systems are also known as section control because they enforce speed over an entire section of a road rather than at fixed locations. Speed enforcement is never popular with motorists, but the important questions are whether section control is effective in reducing accidents, and if they are are more effective than conventional speed enforcement to justify the higher cost.

The main benefit of section control is that it induces more disciplined driving over a long stretch of a road. This is suitable for highways with few entrances and exits, but not much for city streets. Section control is also useful for somewhat longer stretches of road with special conditions, such as bridges or tunnels. Where speeding is a problem on highways, section control does indeed influence driving behaviour in a positive way. Section control is also more difficult to circumvent. With a fixed (possibly known) location, drivers can slow down to avoid being caught speeding, and speed up after they have passed the speeding camera. Some drivers even use radar and lidar detectors to avoid speed traps. None of these evasive tricks work with section control. However, every type of speed enforcement will entice some motorists to evade it. With section control that depends on reading license plates, obscuring license plates will be the evasive approach to tackle.

How effective is section speed enforcement? The European Transport Safety Council reports results from two recent studies in Belgium and Scotland. On the Scottish A9—the road that I have been driving on earlier this summer—excessive speeding of more than 10 mph (16 km/h) has dropped by an impressive 97%, as reported by BBC News on 26 January 2015. Another study by Transport & Mobility Leuven in Belgium found that implementing section control on a stretch of the A10/E40 is effective. The researchers found that:

The is no impact on the capacity of the road [...]. There is a noticeable drop of around 4% in the speeds of cars and vans. More spectacularly, there is a drop of some 25% to 30% in the speed deviations of these vehicles, implying that the traffic stream is more stable due to the trajectory speed control. We also observed a positive impact on traffic safety, with the number of accidents dropping by some 15% (implying 23% less vehicles involved). Furthermore, we estimate a drop of around 29% in light offenders, and an even more pronounced drop of around 78% in heavy offenders.

One of the major benefits of section control is that it homogenizes speed. The difference in speed among vehicles is narrower, and that helps avoid collisions. Encouragingly, the researchers also find that section control does not impede traffic flow. The system is also cost-effective, as the researchers find a 9:1 benefit-cost ratio in their study for the first year (and higher in subsequent years). Another study from Norway also finds that "average speed cameras are significantly more effective [than fixed-point cameras], with a reduction of the driving speed and an associated reduction in injury costs that is up to three times as great."

In British Columbia, speed cameras were a major election issue in 2001, and the new provincial government abandoned photo radar soon after coming into office. Arguably, the cost of implementing photo radar was problematic, and the process of determining where speed enforcement took place was not necessarily driven by accident statistics. However, research on BC's abandoned photo radar program clearly demonstrated societal net benefits despite the program's flaws. Empirical evidence is not always enough to counter perception problems, however. Some people complained that photo radar was merely a "cash grab" to raise revenue. Section control can counter such perception problems. It can be implemented exactly where speeding is a well-documented problem, such as the Sea-to-Sky Highway from Vancouver to Whistler, which even after improvements for the Winter Olympics 2010 remains a hazardous stretch of highway. Because the beginning and end of section control is indicated to motorists clearly through signs, motorists can choose to comply and avoid a fine. Nobody gets "tricked" into a camouflaged speed trap. British Columbia and other provinces in Canada should give section control serious consideration where improving road safety is a real concern. Section control is transparent, easy to understand, and I am optimistic that motorists will come to appreciate it if such a program is administered wisely and fairly.

Section control can also help make variable speed limits more effective. Variable speed signs were introduced on the Sea-to-Sky Highway (at a cost of $12.5m) between Squamish and Whistler, as well as along a stretch of Highway 1 near Revelstoke and on the Coquihalla Highway, this summer. Variable speed limits make good sense when weather conditions change. However, in order to make these speed limits more than mere advisories, section speed control is the right tool to make motorists obey these speed limits.

Some people may object to section control on privacy grounds. Section control requires storing license plate information at several points. What happens to that data? Could it be used for other purposes than speed enforcement? Such concerns can be alleviated through technical and legal safeguards. Data protection can be implemented in real time. Section control entails using two cameras for entry (A) and exit (B). Photos taken at point A are stored until the vehicle passes point B. If a speed violation is detected, the images are stored for issuing a speeding ticket. If no speed violation is detected, both images are immediately deleted. All photos can be encrypted so that only authorized law enforcement personnel can access them. These measures can ensure that section control respects privacy.

Implementing section control in Canada could be complicated by one small problem: some jurisdictions only require rear license plates but not front license plates. Only BC, Mantioba, New Brunswick, and Ontario require license plates on both sides. This means that section control cameras would only be able to rely on reading one side, rather than having cameras facing in both directions to read both front and rear. License plate in European countries have also made use of fonts that are easier to process through optical character recognition. Section control in Canada may require some long-term enhancements to license plate requirements and use.