Is there an unfortunate coincidence to one of the largest commitments to investment in roadside technology being made just at the point when in-vehicle systems, which provide similar outcomes, are starting to become ‘mainstream’? That last term should probably be qualified. Many new vehicles can be bought with features like Adaptive Cruise Control (ACC) and Emergency Automated Braking (EAB), even those towards the affordable end of the market. However, for them to make up the majority of vehicles on the roads it will take at least seven years, but make no mistake, they will undermine the case for traditional solutions. There are also other changes afoot, particularly for Highways England as they implement their new control system CHARM that will provide new ways to connect and communicate with roadside devices. Being presented with this, the first question that many may ask is, “do we continue as planned or do we need to react and adjust our investment plans?” Perhaps we need to look this dilemma in a bit more detail and consider the case for change by focusing on a particular example, as it may be that a compromise is required to help during this transitional period.
MIDAS (Motorway Incident Detection and Automated Signalling) is a well-established, proven and extensively deployed tool used for two key functions, the control and harmonisation of traffic speeds to smooth traffic and the prevention of incidents involving slow-moving or stationary vehicles. It traditionally consists of inductive loops, local processing, a communications network and signalling to provide information to road users. Broadly unchanged over its life, the most significant change has been the migration to side-fire radar as the replacement for inductive loops for determining the prevailing traffic conditions. The future for MIDAS, though, may not be that bright when we consider how these external influences affect its justification.
Let’s start by considering that the evaluation of the performance of MIDAS, to support its business case, was last undertaken over ten years ago and relies on data from the mid-90s onwards. Since then, traffic patterns and volumes, driver behaviour and attitudes, enforcement and prosecution, as well as the application of MIDAS have all changed. The latter is a particularly important point as we have a mixed environment of this system having progressed from post mounted advisory signals widely spaced to intensively deployed lane based signals with mandatory speed limits and now to more widely spaced, combined message signs displaying only a single mandatory speed aspect with advisory text. It is highly likely that all of these factors have influenced the effectiveness of MIDAS and the benefits it provides.
We’ll turn our attention now to the use of inductive loops and, whilst providing highly accurate results when implemented and tuned properly, they present a significant maintenance challenge. This stems from the loops being damaged by wear and tear of the carriageway and its subsequent repair, as well as actually replacing the loops, necessitating lane closures and the exposure of road workers to potential harm. A recent study found that the majority (83%) of MIDAS faults are attributable to loop issues, leading to the adoption of side-fire radar as non-invasive replacement. However, in order to work with MIDAS in its current form the radar heads are connected in a manner where it ‘pretends’ to be a loop through the use of a contact card, which effectively removes access to some of the clever capabilities these devices have. Hold that thought.
If we examine a similar implementation to MIDAS, but more commonly found in our urban centres, we find SCOOT-UTC (Split, Cycle and Offset Optimisation Technique – Urban Traffic Control). This also uses detectors at the roadside, a communications network and signals to inform road users, but the big difference is that the majority of the processing happens centrally and, in more recent times, using a digital communications network similar to that available to MIDAS. Why is this relevant? Mostly, because it proves that a time critical application can use central processing even with geographically remote end devices. CHARM will provide this central processing capability, which would enable the direct connection of both radar and the associated signalling. As CHARM will make use of the NRTS (National Roads Telecommunications System) network, the issue of needing local processing for resilience is addressed and the MIDAS outstation is no longer required.
But wait! Why rely on an end device, providing a point source of monitoring, to determine what operational interventions need to be put in place? Highways England has access to a rich source of data it could use in the form of Floating Vehicle Data (FVD) through its National Traffic Information Service (NTIS), providing much wider network coverage than MIDAS and, therefore, allowing it to monitor parts of the network that don’t currently have any roadside devices, such as trunk roads. This could be directly injected into CHARM at a central location and reduce the need for the deployment and upkeep of a large number of roadside devices and their associated communications links, further reducing costs and maintenance liabilities, as well as reducing street clutter. There are some challenges associated with this related to latency and source density, but some MIDAS functionality could be achieved and some additional functionality introduced.
In going back to our starting point, how do modern vehicles with Advanced Driver Assistance functionality feature in all of this? Well, as previously described, MIDAS offers two key services, namely Controlled Motorways (CM) and Queue Protection (QP). If we consider that:
- Adaptive Cruise Control provides the speed harmonisation outcome that is achieved by CM, then as the proportion of vehicles using this functionality reaches a critical point, which is in the region of 8%; and
- Automated Emergency Braking provides functionality similar to QP, but on a more granular basis.
Then the benefits that this tool provide are quickly eroded. In addition to these features, there are currently connected vehicles trials planned, in the form of UK CITE and A2M2, that will showcase vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communications to facilitate new user services. This will include corporative cruise control, allowing better coordination of vehicle speeds, further enhancing safety benefits, as well as enabling shorter headways that will provide capacity increases. Therefore, it is inevitable that in the near future the continued rollout of one of the most beneficial tools for managing the Strategic Road Network becomes much harder to justify, especially when it is considered that this becomes ‘deployed’ on the whole network by default with the proliferation of equipped vehicles.
Roadside technology may still have some longevity, but the headline conclusion is that now is the time to revisit the business cases for its deployment at a holistic level, as well as at a component level, in order to inform future investment before we have a stockpile of legacy equipment proving a burden rather than benefit. ◆