We’ve previously written about the concerns many in the automotive industry have regarding the safety of smart motorways but a recent report has found that they could be more dangerous than previously thought.
Smart motorways are intended to utilise the space available on the road and technology across the lanes, with smart boards, to allow efficient and timely travel for drivers even in peak traffic. In many places this has meant sacrificing the hard shoulder to allow for an additional lane of traffic and replacing it with designated stopping areas every 1.5 miles and using stationary vehicle detection (SVD) technology.
SVD technology is designed to spot any vehicle which is stationary on the motorway and notify operators so that they can use overhead lanes to move traffic and send emergency services to aid the stranded vehicle. It uses radar, a server and an algorithm to detect the stopped vehicle and then alerts operators at the Highways England regional control centre.
The report published by Highways Magazine suggests the SVD could fail to warn motorway operators of thousands of stranded cars annually, based on analysis of a previous report.
The previous study in 2015 on an eight-mile stretch of the M25 over six days found that 192 stationary vehicle events were detected by the SVD system in place but CCTV later showed that there were 294 events, so more than a hundred we missed giving the system a success rate of just 65%.
62 of the missed events were classed as allowable missed detections and 16 were picked up by cameras but occurred outside the range of the SVD’s radar units at the time, however Highways England has since admitted that their report contained errors.
In 2016 Highway England advised that the report contained 17 events that shouldn’t have been included in the total 294 to begin with and 45 of the allowable missed detections were actually detected by the SVD system but the alert was suppressed. The system allows suppression of alerts so that operators aren’t overwhelmed with lots of unnecessary reports.
Based on this they say the SVD system correctly detected 237 incidents of a total 277 which put it at a success rate of 85.6%, above their required 80% benchmark with the required 3.9% margin of error. They did admit that there were 40 missed events that were genuine and seven of these were linked to blind spots in the radar system.
Though Highways England believes that the errors in the report actually mean that smart motorways are safer and more accurate than previously thought, as well as putting the SVD system above the required threshold the recent analysis conducted by Highways Magazine suggests that the opposite is in fact true.
Their analysis suggests that 2400 stationary vehicles could be missed each year on just an 8 mile stretch of the smart motorway.
With the RAC estimating that there will be 788 miles of smart motorway within the UK motorway network by 2025 this could lead to thousands of dangerous and potentially deadly situations.
Highways Magazine states that Highways England has since admitted to them that a number of these incidents that were classified as detected should have been classified as missed and if they had been then the system would not have met the 80% required threshold.
They criticised the classification of some events as being out of range, as the trial used radar sites that achieved 96.7% coverage of the carriageways, above the benchmark of 90%, and that the remaining 3.3% were not out of range but in blind spots close to the radar devices. Highways Magazine suggests that these blind spots would have been covered by radar devices used and so this implies that the out of range results were due to the limitations of the SVD system.
Another criticism they had was over the suppression of alarms that notified operators of a stationary vehicle. Any stationary vehicle can be a hazard to other drivers and dangerous for occupants still inside the vehicle and so Highways Magazine has highlighted how the suppression of alarms, even manual suppression for maintenance vehicles could be dangerous.
Even when the alerts are issued there are follow up concerns about whether the operators could react in time to prevent a further incident.
A second trial of the SVD system in use on smart motorways was completed in 29018 however Highways England will not publish a report on this as they say it contains sensitive information. They have also said that this trial cannot be compared with others as it was conducted in an area with low, medium and high-flow traffic, whereas the system is primarily designed to operate in low-flow traffic where the motorway is the quietest.
Highways England did advise that the 2018 trial achieved a success rate of 89.6% in low-flow situations and 82.4% for medium and high flow combined.
A Department for Transport spokesperson told Highways Magazine: “Since taking office, the current Secretary of State has expressed his concerns over smart motorways. He has committed £500m to safety improvements and has recently pressed Highways England to further accelerate work in certain areas.”
From 2016 to 2020 there was little movement in the installation of the SVD system across all new and existing all lane running schemes, where there is no hard shoulder, however Transport Secretary Grant Shapps told MPs in January that all retroactive installations would be completed by the end of the year.
Shapps said that the reason behind the failure to install the SVD earlier was that the supplier is a single company and so it was delayed due to this, but he also argued that smart motorways shouldn’t have been opened without the system in place.
A new contract worth £18 million was awarded to the firm Navtech, who supplied the system that was tested previously. A second contract worth £14 million will cover SVD systems on new smart motorway developments. So the system should be in place in on all lane running schemes from January 2022.
We’ll keep you updated as we learn more about smart motorways.
Leave a Comment