The benefits of combining enforcement and traffic management
First publishedin ITS International
Vysionics, which makes the SPECS average speed enforcement system, is looking to expand its systems' capabilities with longer range radar detection and monitoring
Jason Barnes considers how combining enforcement equipment with other traffic management technologies might benefit our future – if only the will were really in place to do so.
During the ITS World Congress in Vienna in October last year, NavTech Radar
and Vysionics ITS announced a strategic partnership that would combine the expertise of Navtech in millimetre-wave wide-area surveillance technology with Vysionics
’ machine vision-based automatic number plate recognition (ANPR) and average speed measurement competencies.
The aim, according to the partners, is to increase the sophistication and utility of existing traffic management and enforcement applications and open the way to new ones – for instance, radar’s longer-range and true all-weather detection capabilities will allow Vysionics’s vision-based systems to ‘see’ much greater distances, enabling much more elaborate and capable enforcement strategies to be designed and implemented, according to Stephen Clark, Navtech’s co-founder and business development director.
Geoff Collins, sales and marketing director of Vysionics, reinforces the point. “ANPR is fantastic at tying down ‘what’ has been done to ‘whom’ but there’s only so far that you can see with it,” he says. “Radar is very good at identifying ‘what’, behaviour, such as illegal manoeuvres, but not necessarily the ‘who’. It is very good at providing prompts for vision-based systems, however.”
Tying the two together is intended to produce far more intelligent solutions than the use of simpler Doppler radar to trigger enforcement cameras. For example, identifying certain behaviours can be identified against not one but a series of thresholds. Illegal manoeuvres could be used as a trigger, with box junction monitoring being a typical application.
Allying previously discrete technologies in order to realise new applications and enrich existing ones blurs the distinctions between various ITS technologies and applications and gives rise to a question: if enforcement specialists were to embrace other ITS technologies, which would they be, and why? There is also the question of what it all means in terms of applications: will we continue to see distinct market verticals or will technological convergence result in operational convergence?
Combining or adding?
In some instances, it is not so much the combining of technologies as the intrusion of a whole new one that will make a difference to enforcement, according to Gatso
’s MD, Timo Gatsonides. Machine vision has been implemented in traffic management in one form or another for quite some years but is finally coming to be recognised as a force for change in itself, he says.
“We’re seeing a lot of progress within the machine vision sector, for applications both inside and outside the traffic sector. A key differentiator is the technology’s versatility – machine vision is so broad that it cannot be pushed into a tight vertical as we might with a loop or radar for detection.”
It follows that we will see new applications emerge. Digitisation has already resulted in convergence between enforcement systems and applications such as presence detection, flow and speed measurement and incident detection, among others. In many places around the world, it is the slow pace of legislative change that has been the obstacle to progress, but Gatsonides sees enforcement applications such as section speed control merging with traffic monitoring, for instance.
He also sees machine vision underpinning a closer relationship between enforcement and road tolling.
“Computer vision readily facilitates applications such as vehicle classification but it also enables, for instance, speed enforcement by vehicle type. In the Netherlands, cars are limited to a top speed of 100km/h but a car with trailer can only legally do 80km/h. You can also monitor dangerous goods vehicles’ use of tunnels and we might finally see a solution which copes adequately with enforcing seat belt use or catches those using mobile phones whilst driving,” he says.
Mention of tolling prompts a discussion of in-vehicle technology. ITS International has reported in the past on the reluctance of many of the connected vehicle research projects to discuss the future application of what they see as safety-related technologies for enforcement purposes. This is despite there being a clear correlation between the two and reflects a concern that such ‘Big Brother’ approaches will discourage consumer up-take.
Gatsonides is convinced that, for the foreseeable future at least, enforcement has to remain outside the vehicle. “Wherever you see an electronic solution implemented, you also see a greater compulsion to tamper with it,” he comments.
“For instance, some people will attempt to use duplicate number plates, but in most countries the incidences are actually quite low when taken across a whole head of population. However when the Netherlands introduced a new smart card for public transport payment, there were hacks posted on the internet within days. If, for instance, we were to use Bluetooth to provide enforcement system pre-emption, I’m sure we’d see more Bluetooth-enabled devices simply being turned off once people became aware of what was happening.
The UK's HADECS3, once deployed, will sit atop an MS4 variable message sign post and could provide traffic data – if the regulatory environment allows (Pic: Highways Agency)
“Any in-vehicle system needs a certified device which is tamperproof,” he continues. “Road user charging requires an on-board unit and GNSS-based technology seems to be the favourite for any nationwide scheme. I’m not suggesting it follows that tolling is a natural route into in-vehicle enforcement, but any government-driven charging scheme is going to want to maximise both revenue and fairness. That implies some robust form of enforcement and I can see an evolution occurring. But it’ll rely in the first instance on one of these national schemes actually progressing to deployment and it will take time.”
Physical vs electronicMassimiliano Cominelli, traffic division manager with vision systems specialist Tattile, sees the addition of new technologies to enforcement systems as belonging more in the electronic than the physical realm. He points to applications that reside within the camera itself, such as ANPR and access control, as examples.
“In South America, for instance, we have customers who are using red light cameras to enforce access restrictions. These incorporate ANPR and work even when the lights are on green,” he says.
“We’ve also developed cameras which allow red light running at night. The signal may be on red as a part of its normal cycle.
If, though, the other carriageway is clear it can make little sense to stop, especially if you’re in a neighbourhood which is not especially safe, but the camera system knows not to generate a violation citation. It’s an example of how applying more intelligence to the camera increases safety even though a ‘traditional’ violation has been committed.”
Tattile is developing multi-lane speed solutions and Cominelli would like to add traffic monitoring capabilities, though these remain paper projects at present. A big factor is price, he says: “People want state of the art but they want it at a low price. The two don’t necessarily go together and in some ways it forces limitations on the technology – and I’m very much of the opinion that technology should be without limits.” The company’s core strength is in machine vision systems for industrial applications, but many of the developments there carry over into the traffic space, Cominelli says.
“A big trend is to have everything done electronically within the camera. For instance, we’ve developed an ANPR solution
which works at up to 250km/h and which is self-triggering, so it doesn’t need an external sensor.
“Perhaps the question to ask is not ‘which technologies can intrude into the enforcement space?’ but ‘how can enforcement intrude more into the traffic management space?’ System evolution means that it’s harder and harder to differentiate by application and market vertical but it also means that consideration can and even should be given to some trade-offs.
“For instance, a vision system might not give the same absolute performance in some circumstances as a loop or laser used for traffic counts – not yet – but it can offer a more than passable capability and, where communities want a non-invasive technology which doesn’t require roads to be dug up for installation, it has to be a realistic choice,” he says. “Similarly, night-time identification of vehicle type by shape might not identify an individual vehicle in the absolute sense, but it does provide a usable vehicle classification solution.”
The same, but differentCarsten Biermann, head of project and product management and marketing at Jenoptik, sees new technology additions as a means of extending systems’ adaptability, rather than a means of introducing new applications. He points to the use of wireless LAN (WLAN) to allow more widespread distribution of a system’s components as an example.
“This would introduce a lot of flexibility in terms of where – above, around and beside the roadside – those components actually reside and has benefits in terms of being able to deploy a system more easily, or else ease operational considerations such as maintenance. There are some type approval/homologation issues, particularly when it comes to ensuring adequate data encryption, but we think the technology is there to do this. Pulling it together is the challenge,” he comments.
“Other areas of interest include the extraction of enforcement images from video instead of still images. Sensor fusion is also something we’re looking at, including the application of light detection and ranging (LIDAR), time-of-flight sensors and wide-angle radar.” Five to ten years out, he says, systems will be much smaller and less energy-hungry, with the flexibility to more widely distribute the individual elements. A sticking point, however, especially in established markets such as Western Europe, will be homologation.
“If you’re going to integrate new technologies into an enforcement camera housing, then you have to make sure that issues such as electro-magnetic interference [EMI], or indeed anything that could possibly affect the enforcement process, are fully solved,” says Biermann.
Retrospective briefMartyn Harriss, enforcement adviser with Redflex, says that to look forward, it’s necessary to look back. He starts before the current economic downturn and the many plans to upgrade older wet film-based enforcement systems which were subsequently put on hold.
“Nevertheless, the EU still has ambitious plans for further reductions in the numbers of KSIs [Killed and Seriously Injured] on its roads. That implies investment in technology and 2013 will, I think, be a turnaround year, with organisations such as Transport for London and others finally looking to replace their old wet-film cameras with more up-to-date digital solutions. The industry has contracted in recent years but we’re still seeing some good stuff coming through the type approval process.”
Stalled plans to replace outdated analogue systems indeveloped markets are starting to come back to life (Pic: TfL)
Harriss expects HADECS3, the third generation of the Highways Agency Digital Enforcement & Compliance System, to break new ground when it passes out of development and into deployment. “This is in line with wider deployment of Managed Motorways, the UK Highways Agency’s take on hard shoulder running,” he says. “We’ll see some new technology elements, such as the use of side-fired dual radars; one as the primary speeding detection system, and a second scanning system which identifies exactly which vehicle is committing the offence and sets it up for an image capture, all whilst removing the need for physical secondary check marks such as road markings or loops.”
A significant feature of HADECS3 is that it is mounted on an MS4 variable message sign pole, rather than an overhead gantry, and is therefore maintainable without having to shut carriageways. Also, by virtue of being able to see across several lanes (during testing, ANPR data has been collected across up to six lanes at night), it has the potential to offer a solution for vehicle count and classification, should the legislative environment evolve to allow it. Further into the future, Harriss sees the possibility of bundling other enforcement applications such as weigh-in-motion in with the systems at the roadside.
He also sees more innovative use of consumer electronics, as is already being explored to an extent in newer enforcement markets such as the Middle East.
Posting a citation to a driver’s home address could become old hat: instead, sending it to a driver’s personal email address or even – using common protocols such as Bluetooth – their mobile phone will become more common.
In the latter case, the ‘flash to bang’ – the time between an offence being committed and notification being transmitted and received – could be measured in tens of seconds.
“It needs the legislative mindset to shift, though,” warns Harriss, “and a driver of that is who exactly enforces. There are three options: the public sector, outsourcing completely to the private sector or insourcing, whereby the public sector staffs systems provided by a private-sector partner such as Redflex.
Service level agreements would be enshrined in contracts and the public sector partner would benefit from there being no capital expenditure.
The private company makes its money from a percentage of the revenues generated from offences.
“There’s risk, in that a successful enforcement strategy actually catches few offenders, because few actually do offend.
But any expert company goes into such an arrangement with its eyes wide open and any business decision carries risk. It’s for the companies with the requisite knowledge to appreciate and manage that risk.”
Legislative concernsMany enforcement experts caveat technological combination progress with comments on legislative concerns – Gatsonides, Biermann and Harriss all make reference to this and it is undoubtedly true that in some countries more than others, the legislative environment continues to stand in the way of any closer integration of enforcement and other traffic management technologies.
More progressive operating paradigms employed in locations such as the Middle East are only slowly filtering back to more established enforcement markets
In the UK, for example, the use of the term ‘safety camera’ rather than ‘speed camera’ is a matter of policy and automated enforcement system operation rests with Safety Camera Partnerships.
For political reasons, a distinction is made and a distance is maintained between traffic operations and enforcement of safer driving.
Peter Hill, director of UK operations at Truvelo, says: “A speed or red light camera installation obviously provides a secure and waterproof enclosure at the side of the road, with a power supply to it. However, this does not mean that it can be crammed full of other traffic technology.
Government departments and police forces are selling the concept to the public that safety cameras are located at accident blackspots with the intention of saving lives and serious injuries. There is a fear that this message might get blurred if other technology shares the site.
“Also, technical issues around electro-magnetic interference come into play,” he says. “At present, the UK Home Office will not permit another piece of kit to be installed in the same enclosure as the speed camera system.
If there was a desire to install an ANPR or a traffic surveillance camera in the same enclosure as a speed camera, that’s certainly achievable but it would perhaps be better to look to do this from the outset and the requisite testing would need to be carried out.”
Contrary thinkingSuch comments raise interesting questions about technology convergence. Politically motivated and artificial separations between traffic enforcement and management may have been implemented with the best of intentions and, in many cases, in order to counter perceptions of enforcement systems as revenue generators, but they stand in the way of good operational and fiscal sense.
Sadly, this is by no means news – digitisation has resulted in enforcement systems that can additionally provide traffic management capabilities being available to the market for some years.
The Navtech Radar/Vysionics tie-up provides an example of how traffic management and enforcement technologies can be brought together but the expected result – at this stage, at least – is enhanced enforcement, not the addition of, or combination with, traffic management.
Enforcement systems manufacturers, being in the business of selling hardware, are keen to add functionality and increase their sales prospects but remain hamstrung, especially in more established markets. In some cases, it is not too much a stretch of the imagination to see enforcement systems offering traffic management capabilities which match the best-in-class of other single-use technologies. This is especially so with the application of machine vision. But even where that is not the case, there is still room for enforcement and traffic management to be combined. Tattile’s Massimiliano Cominelli makes a valid point when he says that a capability may not be the capability but can be more than adequate in some circumstances.
The ITS industry has a hard enough time, in the face of entrenched thinking and vested interest, persuading decision-makers of its potential in terms of cost savings and added value. And this gives rise to a paradox: on the one hand, systems manufacturers are being asked to increase their products’ utility by adding features that theyare then pretty much obliged to write off - if not at a loss, exactly, then certainly as inherent greater ‘value’ which then retails for no greater cost. On the other hand, legislative change takes place at a glacial rate.
This focus on product, rather than process, is increasingly absurd at a time when so much emphasis is placed on saving money and – but for some changes to the rule books – one system could so easily take on the role of several.