National Highways is not deploying video technology that could save lives on ‘smart motorways’ after it was told its CCTV systems are not up to scratch.
The government-owned company has successfully tested a CCTV-based video analytics system, which can identify stranded vehicles in live lanes in all conditions.
AA president Edmund King has expressed surprise that the technology, which he said could be ‘incredibly useful’, is not being taken forward.
As Highways has reported, in 2020 Highways England (as it was known) carried out a large-scale trial of a system that analyses CCTV images to detect stopped vehicles.
It subsequently declined to state the outcome of the trial despite the technology being cited in the Government’s smart motorway action plan as ‘providing another option alongside current radar technology’
However, a report published last year, produced by consultants on behalf of strategic roads monitor the Office of Rail and Road, disclosed that the video analytics service platform (VASP) in the trial delivered ‘alerts for incidents of interest’ to operators’ desks, allowing quicker response.
Two types of high priority incident– stationary vehicles and congestion – were detected. The report stated: ‘For the two high priority capabilities tested, the trial reported results which indicated a high level of precision.’
However, it also noted that the analytics system is limited by the CCTV cameras not having complete automatic coverage of the network as they are set to point one way and have to be reoriented manually.
It stated that for this reason, CCTV-based analytics is always likely to be secondary to the existing radar-based stopped vehicle detection (SVD) system.
The report added: ‘Although external to the VASP system, the trial did observe external factors that could limit the use of CCTV to detect stopping vehicles, the main one being the reliability of camera feeds.’
It also noted that Highways England planned to launch a second trial to test the results in winter weather and light at the same location.
A National Highways spokesperson has now told Highways that the second trial was carried out over a six-month period through the winter of 2020/21, with additional locations included to ensure that all expected street lighting conditions were also validated in winter conditions.
The spokesperson added: ‘The performance of the analytics was very slightly reduced by the winter lighting and weather, but was shown to be able to reliably detect events during adverse winter conditions.’
National Highways’ two-year update on its progress against the Government’s 2020 smart motorway ‘stocktake and action plan’, published in May this year, noted that the CCTV analytics trial had been completed but did not disclose that this had involved a second trial or that both had been successful.
The company did not explain why it is not actively taking forward the technology. Its spokesperson said: ‘Our preferred approach is the continued rollout of stopped vehicle detection. It is the most effective solution.
‘We have delivered on our commitment to put radar-based SVD technology in place on all existing ALR motorways. SVD is also being added to ALR schemes still under construction, with no new stretches opening without it. The technology is capable of detecting stopped vehicles in live lanes, automatically alerting drivers with a report of obstruction signs, and sending alerts to control rooms. The alerts are then investigated by operators who can then close lanes with a Red X and set speed limits.
‘In line with our company approach for continuous improvement, we will continue to review and, where possible, improve how we identify stopped vehicles.’
Despite National Highways’ description of SVD as ‘the most effective solution’, its assertion that it is ‘capable’ of detecting stopped vehicles is noticeably weaker than the description of the new system as providing ‘a high level of precision’ and ‘able to reliably detect events’ in poor conditions.
The company has set a benchmark for the existing SVD system of detecting just 80% of stopped vehicle incidents. Last year, Highways reported that it only sent alerts for 65% of incidents in trials and only achieved the 80% benchmark after a number of missed detections were excluded from the calculations.
AA president Edmund King told Highways: ‘Smart motorway technology, coupled with human interventions, are crucial to keeping drivers and passengers safe should they break down in a live lane. We pushed for the rollout of SVD and don't believe any stretches of motorway should have opened without it.
‘But as we have experienced with recent glitches in the system, sometimes it helps to have a belt and braces approach. We believe CCTV could be incredibly useful as a backup system to SVD to help identify broken down vehicles.’
The trial followed concerns over the effectiveness of the existing SVD system, which National Highways’ chief engineer, Mike Wilson, described in a 2019 letter to a coroner as ‘a radar system which detects stopped vehicles in low flow conditions’.
Mr Wilson added: ‘We are exploring other technologies which could reduce the risk to stopped vehicles in higher flow conditions.’
It now appears that National Highways has identified that CCTV-based technology would reduce the risk to stopped vehicles but, for reasons that are unclear, has chosen not to use it.