Cotares Limited

Traffic

Map representation

Our core map break the road network down into uni-directional stretches of road. Thus a two-way single-carriageway road will be represented by two "Links", one for each direction. A one-way single carriageway road will be represented by a single "Link".

The connectivity between these sections of road are represented by "NextLinks". If a vehicle can drive directly from Link A to Link B, there will be a NextLink AB.

During route computations, we consider a vehicle to lie midway along one of the Links. The NextLinks indicate which other sections of road the vehicle can directly reach, and the time, distance, fuel etc. required to get there. In this way, we directly encompass the different traffic conditions or speed profiles that may be encountered when queueing to turn left, right or proceed ahead.

Furthermore, when performing map matching tasks, perhaps to extract vehicle speeds from floating vehicle data, we use the middle of Links as the reference timing points. For a vehicle that moves along Link A then Link B, we would extract the time taken from the middle of Link A to the middle of Link B, and store this as the time to traverse the NextLink AB. This automatically extracts all the dependencies on choice of turns, and is also less noisy because vehicles tend to move more smoothly across the middle sections of Links than at their ends.

When U-turns are permitted, we simply include a NextLink between the two directional Links of a road. When there are time-dependent turn restrictions, we can apply them to the NextLinks. In Europe, there are around 120,000 road segments or turns that have "Validity Periods" attached, and we model them all precisely.

Complex prohibited manoeuvres, multiple lanes etc. are handled by duplicating sets of Links as "Virtual Links" whose NextLinks reflect the driving restrictions and the different speeds, with no need for the router itself to take special account of this.

Traffic congestion handling

Historical and real-time updates are fully integrated.

Batch updates of traffic speeds are tagged so that different countries or suppliers can be independently updated.

We support a wide variety of Map-Independent Location-Referencing formats including TMC codes, Agora-C, and dynamic matching of LatLon trails.

Traffic speed profiles can specify absolute times in the future or past, and can be independently specified for each turn through an intersection (see Map Representation above).

If you are a traffic data provider, our systems are also ready to accept:

  • Detailed speed profiles describing how queues are expected to grow and dissipate.
  • Expected traffic volumes to help us to choose avoiding routes.
  • Confidence and Range values to optimise mean or worst-case journey times.

We also separate traffic information into high and low priorities so that, for example, closure information can override real-time speed data, or vice versa.

Local and Global routing alternatives

When severe congestion or a closure affect a major artery, most routers will pick another major route, or just re-time the artery. This is because they have already ruled out the use of slower roads through hierarchical (Functional Class based) methods. Our router is capable of selecting both local diversions around closures and global alternative routes so that the driver can choose.

If traffic volumes are low, perhaps at midnight, the driver may choose to go near the incident (here shown interrupting the red route) and follow a local avoiding route:

Local diversion through Cambridge

If traffic volumes are high, perhaps in a peak period, they may choose to avoid the area altogether because nearby roads will become congested as the traffic avoids the incident, choosing the route via Ely below:

Choices include more distant route

The presentation of Choice Routes™ allows us to inform the driver about likely journey times, fuel costs, types of road etc., but leaves them to make the final decision. This avoids "deskilling" the driver, and allows them to use their own greater knowledge of time constraints, risk, familiarity etc. to choose the best route for them.