Hybrid detection technologies are set to become a major trend, according to Iteris’s Bill Sowell
Jason Barnes looks at where the detection and monitoring sector is heading.
In the future, there will be no such thing as an un-instrumented road. Just a short time ago, that could have been a quote from a high-level policy document but with the first arrivals of vehicles with 802.11p connectivity – the door-opener to Vehicle-to-X (V2X) applications – it’s a statement which has increasing validity. The technology which uses our roads will also provide information on road conditions but V2X isn’t the only game in town.
Until recently the inductive loop was the only real choice, and then along came video, allowing traffic controllers to ‘see’ in the literal sense. More recently, innovations in technologies such as radar have shown that sometimes it is better without sight – an observation which holds equally true in the case of crowdsourced data using smart devices.
Wider ITS trends such as the desire to retain the integrity of the road surface wherever possible by using non-intrusive solutions, the need to be cost-effective and the arrival of new neater and sometimes simpler technologies, are all influencing the sector. So are authorities facing massive redundancy in terms of the systems deployed or massive redundancy in terms of the mutually supporting data sets which a plethora of complementing technologies can offer?
557 Clearview Traffic, sees aspiration coming up against hard reality with timescales for the next-generation detection and monitoring systems. Although V2X systems are expected to eventually take over from external detection and intrusive solutions such as inductive loops, there’s a need to guarantee statistically meaningful representations of current road conditions. That’s promoting moves into mobile phone-based solutions in the interim, but for certain high-accuracy applications the inductive loop still reign supreme, he says.
“Clearview is providing and maintaining the vehicle data capabilities on 14 of the 15 DBFO [Design-Build-Finance-Operate] projects underway currently in the UK. These operate shadow tolling regimes and classification has to be reliable and accurate because vehicle numbers and types dictate what an operator is paid. We’re currently achieving 99.6% accuracy using inductive loops.”
Loops are best where large and detailed or highly accurate classification is critical but for other applications there are credible alternatives, Muspratt continues. Clearview brought magnetometer technology to the UK traffic management market in 2009 and it has enjoyed relatively rapid uptake.
“There’s a continuing belief that the loop is 100% accurate and any other technology has to prove itself against it – that if a loop shows one thing and the new technology another, the latter must be wrong. That’s not true. Take the example of an intersection as the lights turn green. The first four or five cars will move off pretty much together and a loop will often detect that as a single large vehicle, whereas a magnetometer will provide an accurate count. In that instance, the new technology is correct.”
In his experience, roadside technologies such as side-fired radar hold promise but struggle because of obscuration in multi-lane environments – and the need for an 8-10m mast increases installation costs. Magnetometers, although in-road, are far less intrusive, needing only a 100x50mm hole as wireless operation eliminate ducting and trenching. At a four-way MOVA (microprocessor optimised vehicle actuation) -equipped intersection, and with the ability to offer inward detection out to 80-100m, magnetometers offer a significant capital cost saving and are good for 10 years.
UK-wide, Clearview now has over 500 junctions in operation using the technology.1466 Transport for London (TfL) alone is using magnetometers in conjunction with split cycle offset optimisation (SCOOT) at 200 intersections.
Muspratt: “Our magnetometers can be installed without having to close whole roads – it can be done lane by lane in just 15 minutes. They were tested against loops and side-fired radar before being selected. Above-ground radar proved very good for stop-line detection but still required power cabling, so the capital savings weren’t as great as might be expected.
“TfL is installing low-level signals and implementing cyclist-friendly signal timing regimes. We’re looking at using the same wireless protocol but with in-ground radar which sits in a similarly sized core-drilled stud. There’s a need to detect both the presence and absence of cyclists where new signal phasing is implemented. Timings are altered to allow cyclists a faster getaway from red or to protect them from accidents involving turning lorries but if there are no cyclists it makes no sense to extend timings as intersection efficiency will be reduced. Upward-facing radar was chosen because more and more bikes have less metal content and where the separation between a cycle and live lanes may only be a white line you need its accuracy.”
Clearview is considering using in-ground radar for pedestrian detection, extending coverage to look at the kerbside. There is, Muspratt says, considerable scope for cost-benefit savings based on having a fully wireless network. Bay parking is another area where the company is enjoying success, as authorities look to provide notice of available capacity via online and smart device applications.
163 Inrix’s vice president business development, EMEA, sees the big shifts being in infrastructure and data ownership: “There’s an increasing recognition that private-sector data and traffic information providers are of growing value. The days have gone when authorities wanted total control of all of their sources of information.”
Inrix’s GPS-based, crowdsourced, aggregated data services were used during London’s Summer Olympics and the Golden Jubilee and are still used for events like the New Year celebrations and London Marathon. In many respects, the concept is similar to that of floating vehicle data, although technological advances and the arrival of GPS-equipped smart devices have facilitated a move from strategic routes into more complex urban settings.
Woolard sees great potential in non-fixed data systems; nevertheless, he still feels that a multi-source approach is best. “The quantum shift is from something like ‘here’s a loop-based SCATS system’ to ‘here’s something which uses loops, vision and GPS’,” he explains.
“We’ve been extending our data to new applications, not just real-time traffic. Throughout 2013 we undertook projects monitoring the ‘movement’ of people, specifically for the urban environment. This builds on the work we did during the Olympics. Post-2012 we’ve continued to work with TfL and others to develop our technology for use in a multimodal context, for instance to monitor transport entry points, and we’ve been getting a lot of interest.
“‘Crowd analytics’ is a misleading term. We’re looking to shift the focus to ‘population or data analytics’ because the technology’s not tied to specific events but to an overview of transport movements. We’re not at the level yet where we can tell an individual bus that there are six or 20 people at a particular stop but that has to be the goal.”
Inrix is already a connected vehicle technology supplier, supplying dynamically updated traffic information and driver services to many OEMs.
Woolard: “There’ll continue to be distinctions between what we do and the safety-related V2X applications. They’ll need a 5.9GHz-type technology – you’re not going to do the near-zero-latency stuff with something like cellular. It comes back to having that mix; cellular data has the advantage of huge sample sizes but will it replace fixed infrastructure for traffic management? Not yet, and maybe never.
“We’ve being doing some Origin-Destination [O-D] work using the technology. The use of loops and then Automatic Number Plate Recognition [ANPR] for O-D studies has recently been challenged by the arrival of Bluetooth-based solutions which by comparison with ANPR are less costly to install and require rather smaller sample sizes to be effective. Our cellular solution, using highly aggregated data which doesn’t encroach on privacy, competes with or can complement both.”
819 NavTech Radar specialises in solutions for longer-range detection and monitoring and company co-founder Stephen Clark sees the wider adoption of Hard Shoulder Running (HSR) as a major opportunity.
“HSR works. It’s increasingly seen as a means of increasing capacity quickly and at relatively little expense but initial deployments, such as the M42 near Birmingham, England, have been technology-heavy. There’s a desire to reduce the technology content.
“Road network operators want extra capacity but they want it to be safe. There’s been a shift in operational thinking to a situation where hard shoulders remain open constantly and the aim, in the case of a three-lane highway with a hard shoulder is that ‘four’ should be at least as safe as ‘three plus one’. But while that four-lane concept is safer overall a single stopped vehicle presents a greater risk. That makes near-instant incident detection essential.
“Solutions such as eCall may take relatively little time to inform first responders of a problem but that’s still a long while in live lanes at motorway speeds. Connected vehicles are exciting and offer potential safety improvements but there’ll still be the issues of unequipped, legacy vehicles or power or system failure in the case of an equipped vehicle. Fast, reliable detection and monitoring from the infrastructure side isn’t going away just yet.”
Navtech recently launched a new generation of commercial-grade radar for traffic applications. The all-new design has resulted in a smaller and lighter radar head with less wind resistance. Software commissioning tools have been enhanced and it offers greater mounting flexibility while retaining the ability to detect a vehicle-sized object at over 1km and an adult human at 700m.
“Our solutions providers want to deliver more than just velocity and distance – information like the angle to an oncoming or stationary object is also desired. There’s interest in the ability to use a single device to monitor multiple lanes or crossroads with several approaches. We’re also being asked to monitor over greater distances and wider areas. Such monitoring delivers a history relating to every tracked object which is useful for traffic light control as well as highway applications such as count and classification.”
Smart, multi-use devices with a high level of integration are also a significant trend, Bömmel continues. He cites examples such as traffic signals which also trigger red light and speed enforcement systems.
“It’s increasingly important to deliver solutions that offer easy mounting and integration with other infrastructure. But creating a device, as we have done, that can be used for different applications like traffic light control, speed, red light and wrong turning enforcement, counting and classifying, monitoring of emergency lanes and so on, means developing a solution that can gather as much information as possible.”
Competitors allude to radar’s difficulties in certain situations, including obscuration. InnoSenT addresses the issues with some novel antenna arrays and by using multiple antennae to provide area coverage.
Bömmel says ‘old-fashioned’ antenna approaches could not meet the market’s requirements and smarter concepts were necessary. “So InnoSenT has pursued a series of antenna concepts including MIMO [multiple-input/multiple-output], digital beam forming, phased arrays and phase comparison architectures. Such concepts can only be realised through close interaction with the signal processing aspects of a system and sophisticated designs.”
The results, he says, satisfy the most sophisticated applications including the detection and monitoring of vulnerable road users in complicated circumstances such as an injured or immobile person on a crossing or in the live lanes. This is far more sophisticated than just detecting movement and requires very fast and accurate positioning capabilities.
The company already has an eye on systems for traffic monitoring in the Vehicle-to-X and autonomous driving environments.
“Future traffic monitoring systems will actively prevent accidents by providing communication channels between infrastructure and vehicles. Most areas will be covered by traffic monitoring systems and drivers will be actively informed of traffic occupancy, road and environmental conditions. This can only be generated by using the next generation of traffic monitoring systems and depending on the application, I expect to see both pure radar systems as well as fusion with vision, lidar and time-of-flight cameras,” says Bömmel.
148 Wavetronix, says there’s also a need to continue developing single solutions to their fullest extent. The company’s latest SmartSensor HD epitomises that thinking, he says, essentially offering ‘more of everything’ over previous incarnations.
“We’re concentrating on that higher standard of data - it’s something customers are willing to pay for. Development has been a two-way process – the industry’s request for better data and our response with improved technologies that allow data processing to a greater accuracy than was previously available.”
‘Better data’ is a vague term but, like Bömmel, Rose highlights feature-rich data in relation to better positional and speed information, vehicle separation and classification.
“Essentially, we’re saying, ‘Not only was a vehicle detected, but here are all the related stats.’ Two to five years out, we will see a fundamental push to adopt technologies which increase safety and reduce congestion. Our UK personnel have been working with the authorities on the Aiming for Zero strategy to reduce casualties among in-road workers. We’re now seeing similar zero-death initiatives coming through in the US. Many US states have initiated similar campaigns and I expect to see a lot of bipartisan support for infrastructure funding that encourages out-of-the-road systems which promote safer conditions for both drivers and road workers.”
Data, he says, is addictive: “There’s a need for data in a real-time context but it also feeds the public debate on where transportation is heading in the longer term. Accurate vehicle detection will remain the cornerstone of that decision-making.”
He also predicts interesting things happening in emerging economies. “There’s going to be an explosion of interest in improving transportation management in developing regions, and there’s continuing2000 World Bank investment in areas like the north-west African nations and India. These markets are trying to facilitate cost-effective routes to commercialisation and there’s a system- rather than device-led approach which features multiple elements. But a strong foundation of data is the key.”
“Counts, speeds, occupancies and headways are all aggregated and then fed to a variety of users. The important part to note is that we’ve moved to a situation of predictive traffic modelling based on the large amounts of data that we now gather and process.
“Connected vehicles and hybrid detection are clear trends. North America might be behind Europe and Japan at the moment but there’s big excitement over the related information packets defined in the recent NHTSA ruling. Location, speed and direction of travel are all of interest to those in intersection management– 15-20 years out, the intersection sensor of choice will be the vehicle itself.”
Older-generation technologies including loops will still be around, he feels, especially as the expense of moving to newer solutions may not be so justifiable in less densely populated areas. “Why go to the expense if there’s no pressing need? The question for cooperative systems is, ‘What’s the adoption rate going to be?’,” Sowell adds.
Prominent among local jurisdictions ‘asks’ is bicycle detection. In the US, California started the trend by requiring that all new or refurbished intersections must be bicycle-friendly. Iteris is supplying a video-based solution which provides cycle detection and differentiation.
“We now have 25 cities in the US which are officially bicycle-friendly and a further 50 who want to be so,” Sowell continues. “Another growth area is the provision of iPad and tablet apps which allow traffic engineers to remotely view data – including real-time video. Loops, video, infrared and radar are all reliable when well-maintained but that can’t always be taken for granted in the current financial climate, so user-friendly and remote set-up and access adds substantial value.”
In the future, there will be no such thing as an un-instrumented road. Just a short time ago, that could have been a quote from a high-level policy document but with the first arrivals of vehicles with 802.11p connectivity – the door-opener to Vehicle-to-X (V2X) applications – it’s a statement which has increasing validity. The technology which uses our roads will also provide information on road conditions but V2X isn’t the only game in town.
Until recently the inductive loop was the only real choice, and then along came video, allowing traffic controllers to ‘see’ in the literal sense. More recently, innovations in technologies such as radar have shown that sometimes it is better without sight – an observation which holds equally true in the case of crowdsourced data using smart devices.
Wider ITS trends such as the desire to retain the integrity of the road surface wherever possible by using non-intrusive solutions, the need to be cost-effective and the arrival of new neater and sometimes simpler technologies, are all influencing the sector. So are authorities facing massive redundancy in terms of the systems deployed or massive redundancy in terms of the mutually supporting data sets which a plethora of complementing technologies can offer?
Realistic timescales
Graham Muspratt, Group Product Manager with“Clearview is providing and maintaining the vehicle data capabilities on 14 of the 15 DBFO [Design-Build-Finance-Operate] projects underway currently in the UK. These operate shadow tolling regimes and classification has to be reliable and accurate because vehicle numbers and types dictate what an operator is paid. We’re currently achieving 99.6% accuracy using inductive loops.”
Loops are best where large and detailed or highly accurate classification is critical but for other applications there are credible alternatives, Muspratt continues. Clearview brought magnetometer technology to the UK traffic management market in 2009 and it has enjoyed relatively rapid uptake.
“There’s a continuing belief that the loop is 100% accurate and any other technology has to prove itself against it – that if a loop shows one thing and the new technology another, the latter must be wrong. That’s not true. Take the example of an intersection as the lights turn green. The first four or five cars will move off pretty much together and a loop will often detect that as a single large vehicle, whereas a magnetometer will provide an accurate count. In that instance, the new technology is correct.”
In his experience, roadside technologies such as side-fired radar hold promise but struggle because of obscuration in multi-lane environments – and the need for an 8-10m mast increases installation costs. Magnetometers, although in-road, are far less intrusive, needing only a 100x50mm hole as wireless operation eliminate ducting and trenching. At a four-way MOVA (microprocessor optimised vehicle actuation) -equipped intersection, and with the ability to offer inward detection out to 80-100m, magnetometers offer a significant capital cost saving and are good for 10 years.
UK-wide, Clearview now has over 500 junctions in operation using the technology.
Muspratt: “Our magnetometers can be installed without having to close whole roads – it can be done lane by lane in just 15 minutes. They were tested against loops and side-fired radar before being selected. Above-ground radar proved very good for stop-line detection but still required power cabling, so the capital savings weren’t as great as might be expected.
“TfL is installing low-level signals and implementing cyclist-friendly signal timing regimes. We’re looking at using the same wireless protocol but with in-ground radar which sits in a similarly sized core-drilled stud. There’s a need to detect both the presence and absence of cyclists where new signal phasing is implemented. Timings are altered to allow cyclists a faster getaway from red or to protect them from accidents involving turning lorries but if there are no cyclists it makes no sense to extend timings as intersection efficiency will be reduced. Upward-facing radar was chosen because more and more bikes have less metal content and where the separation between a cycle and live lanes may only be a white line you need its accuracy.”
Clearview is considering using in-ground radar for pedestrian detection, extending coverage to look at the kerbside. There is, Muspratt says, considerable scope for cost-benefit savings based on having a fully wireless network. Bay parking is another area where the company is enjoying success, as authorities look to provide notice of available capacity via online and smart device applications.
One careful owner…
Danny Woolard,Inrix’s GPS-based, crowdsourced, aggregated data services were used during London’s Summer Olympics and the Golden Jubilee and are still used for events like the New Year celebrations and London Marathon. In many respects, the concept is similar to that of floating vehicle data, although technological advances and the arrival of GPS-equipped smart devices have facilitated a move from strategic routes into more complex urban settings.
Woolard sees great potential in non-fixed data systems; nevertheless, he still feels that a multi-source approach is best. “The quantum shift is from something like ‘here’s a loop-based SCATS system’ to ‘here’s something which uses loops, vision and GPS’,” he explains.
“We’ve been extending our data to new applications, not just real-time traffic. Throughout 2013 we undertook projects monitoring the ‘movement’ of people, specifically for the urban environment. This builds on the work we did during the Olympics. Post-2012 we’ve continued to work with TfL and others to develop our technology for use in a multimodal context, for instance to monitor transport entry points, and we’ve been getting a lot of interest.
“‘Crowd analytics’ is a misleading term. We’re looking to shift the focus to ‘population or data analytics’ because the technology’s not tied to specific events but to an overview of transport movements. We’re not at the level yet where we can tell an individual bus that there are six or 20 people at a particular stop but that has to be the goal.”
Inrix is already a connected vehicle technology supplier, supplying dynamically updated traffic information and driver services to many OEMs.
Woolard: “There’ll continue to be distinctions between what we do and the safety-related V2X applications. They’ll need a 5.9GHz-type technology – you’re not going to do the near-zero-latency stuff with something like cellular. It comes back to having that mix; cellular data has the advantage of huge sample sizes but will it replace fixed infrastructure for traffic management? Not yet, and maybe never.
“We’ve being doing some Origin-Destination [O-D] work using the technology. The use of loops and then Automatic Number Plate Recognition [ANPR] for O-D studies has recently been challenged by the arrival of Bluetooth-based solutions which by comparison with ANPR are less costly to install and require rather smaller sample sizes to be effective. Our cellular solution, using highly aggregated data which doesn’t encroach on privacy, competes with or can complement both.”
Radar developments
Radar continues to make its presence felt with new and maturing applications. UK-based“HSR works. It’s increasingly seen as a means of increasing capacity quickly and at relatively little expense but initial deployments, such as the M42 near Birmingham, England, have been technology-heavy. There’s a desire to reduce the technology content.
“Road network operators want extra capacity but they want it to be safe. There’s been a shift in operational thinking to a situation where hard shoulders remain open constantly and the aim, in the case of a three-lane highway with a hard shoulder is that ‘four’ should be at least as safe as ‘three plus one’. But while that four-lane concept is safer overall a single stopped vehicle presents a greater risk. That makes near-instant incident detection essential.
“Solutions such as eCall may take relatively little time to inform first responders of a problem but that’s still a long while in live lanes at motorway speeds. Connected vehicles are exciting and offer potential safety improvements but there’ll still be the issues of unequipped, legacy vehicles or power or system failure in the case of an equipped vehicle. Fast, reliable detection and monitoring from the infrastructure side isn’t going away just yet.”
Navtech recently launched a new generation of commercial-grade radar for traffic applications. The all-new design has resulted in a smaller and lighter radar head with less wind resistance. Software commissioning tools have been enhanced and it offers greater mounting flexibility while retaining the ability to detect a vehicle-sized object at over 1km and an adult human at 700m.
Radar at intersections
With both automotive and traffic management sector business units, InnoSenT is able to cross-pollinate the two with radar-based technology operating at 24GHz, says Christian Bömmel, the company’s head of sales. InnoSenT supplies technology to several traffic management solutions providers and he discerns a significant trend towards the use of 2D radar.“Our solutions providers want to deliver more than just velocity and distance – information like the angle to an oncoming or stationary object is also desired. There’s interest in the ability to use a single device to monitor multiple lanes or crossroads with several approaches. We’re also being asked to monitor over greater distances and wider areas. Such monitoring delivers a history relating to every tracked object which is useful for traffic light control as well as highway applications such as count and classification.”
Smart, multi-use devices with a high level of integration are also a significant trend, Bömmel continues. He cites examples such as traffic signals which also trigger red light and speed enforcement systems.
“It’s increasingly important to deliver solutions that offer easy mounting and integration with other infrastructure. But creating a device, as we have done, that can be used for different applications like traffic light control, speed, red light and wrong turning enforcement, counting and classifying, monitoring of emergency lanes and so on, means developing a solution that can gather as much information as possible.”
Competitors allude to radar’s difficulties in certain situations, including obscuration. InnoSenT addresses the issues with some novel antenna arrays and by using multiple antennae to provide area coverage.
Bömmel says ‘old-fashioned’ antenna approaches could not meet the market’s requirements and smarter concepts were necessary. “So InnoSenT has pursued a series of antenna concepts including MIMO [multiple-input/multiple-output], digital beam forming, phased arrays and phase comparison architectures. Such concepts can only be realised through close interaction with the signal processing aspects of a system and sophisticated designs.”
The results, he says, satisfy the most sophisticated applications including the detection and monitoring of vulnerable road users in complicated circumstances such as an injured or immobile person on a crossing or in the live lanes. This is far more sophisticated than just detecting movement and requires very fast and accurate positioning capabilities.
The company already has an eye on systems for traffic monitoring in the Vehicle-to-X and autonomous driving environments.
“Future traffic monitoring systems will actively prevent accidents by providing communication channels between infrastructure and vehicles. Most areas will be covered by traffic monitoring systems and drivers will be actively informed of traffic occupancy, road and environmental conditions. This can only be generated by using the next generation of traffic monitoring systems and depending on the application, I expect to see both pure radar systems as well as fusion with vision, lidar and time-of-flight cameras,” says Bömmel.
Pure solutions
While multi-use technologies and system fusion are fashionable, Mike Rose, chief business development officer with“We’re concentrating on that higher standard of data - it’s something customers are willing to pay for. Development has been a two-way process – the industry’s request for better data and our response with improved technologies that allow data processing to a greater accuracy than was previously available.”
‘Better data’ is a vague term but, like Bömmel, Rose highlights feature-rich data in relation to better positional and speed information, vehicle separation and classification.
“Essentially, we’re saying, ‘Not only was a vehicle detected, but here are all the related stats.’ Two to five years out, we will see a fundamental push to adopt technologies which increase safety and reduce congestion. Our UK personnel have been working with the authorities on the Aiming for Zero strategy to reduce casualties among in-road workers. We’re now seeing similar zero-death initiatives coming through in the US. Many US states have initiated similar campaigns and I expect to see a lot of bipartisan support for infrastructure funding that encourages out-of-the-road systems which promote safer conditions for both drivers and road workers.”
Data, he says, is addictive: “There’s a need for data in a real-time context but it also feeds the public debate on where transportation is heading in the longer term. Accurate vehicle detection will remain the cornerstone of that decision-making.”
He also predicts interesting things happening in emerging economies. “There’s going to be an explosion of interest in improving transportation management in developing regions, and there’s continuing
Conditional funding
Bill Sowell, Iteris’s vice president of sales and marketing, says with MAP-21 funding being conditional on projects meeting measures of effectiveness, the desire for data has never been greater. ‘Big Data’ is already here and being used by the likes of Iteris on its contract with Caltrans to aggregate California’s road sensor data cross-state and across various sensor types.“Counts, speeds, occupancies and headways are all aggregated and then fed to a variety of users. The important part to note is that we’ve moved to a situation of predictive traffic modelling based on the large amounts of data that we now gather and process.
“Connected vehicles and hybrid detection are clear trends. North America might be behind Europe and Japan at the moment but there’s big excitement over the related information packets defined in the recent NHTSA ruling. Location, speed and direction of travel are all of interest to those in intersection management– 15-20 years out, the intersection sensor of choice will be the vehicle itself.”
Older-generation technologies including loops will still be around, he feels, especially as the expense of moving to newer solutions may not be so justifiable in less densely populated areas. “Why go to the expense if there’s no pressing need? The question for cooperative systems is, ‘What’s the adoption rate going to be?’,” Sowell adds.
Prominent among local jurisdictions ‘asks’ is bicycle detection. In the US, California started the trend by requiring that all new or refurbished intersections must be bicycle-friendly. Iteris is supplying a video-based solution which provides cycle detection and differentiation.
“We now have 25 cities in the US which are officially bicycle-friendly and a further 50 who want to be so,” Sowell continues. “Another growth area is the provision of iPad and tablet apps which allow traffic engineers to remotely view data – including real-time video. Loops, video, infrared and radar are all reliable when well-maintained but that can’t always be taken for granted in the current financial climate, so user-friendly and remote set-up and access adds substantial value.”
