Location data provides valuable information to traffic planners and engineers, allowing for the collection of metrics such as origin-destination, speed, and travel time data. Location data is collected from moving vehicles on the road, and can be collected from different connected devices. It plays a primary role in how big data traffic analytics operate, as accurate, relevant data is the foundation of traffic studies. Different types of location data have many similarities, which can at times make them hard to distinguish. However, there are some major differences depending on the categorization of the data. Choosing the right location data is vital to meeting your project goals. 

What is Location Data?

Typically, location data gets broken down into two distinct categories: Data coming from location-based services apps and connected vehicle data. The data from location-based services apps come from mobile phones, transmitting data as a motorist travels, but it does not track the actual vehicle. Connected car data, on the other hand, is data accessed directly from cars equipped with connected vehicle technology. Data is sent to wireless networks, which allow vehicles to communicate bi-directionally with other vehicles or infrastructure. This communication allows vehicles to share information regarding their location, speed, and other parameters in real-time with the manufacturer or fleet centers. Within these two data categories, the data can be further classified based on update frequency, either high or low, and location accuracy.  

Location-Based Services Apps  

Location data coming from location-based services apps provide useable data, however, this data has its limitations compared to connected vehicle data. Since data is coming from mobile phones, data risks being captured in duplicates if there are multiple connected phones inside a vehicle. The metrics that stem from this data are very basic, as the data received cannot properly estimate various factors, such as precise travel path, along with many extraneous factors influencing the data received. This is due to an inaccuracy in location, low ping frequency, and the inability to distinguish multiple devices from the same car. Here, some of the key differences between low and high frequency data: 

Low Frequency

This data updates every 1-2 hours and is sourced from mobile apps. It cannot determine exact trajectories, nor can it complete origin-destination analysis, because the data does not capture exact speeds or travel time with low frequency location-based services apps. Update latency is every 6 hours, leading to a delay in reception of the data and delaying potential actions to be taken. This means there is no ability to view congestion, incidents, or blockage in real-time.  

Given the lack of accurate information provided from this type of data, many analyses cannot be done with just low frequency location-based services app data. While it can be used for specific cases, it is the lowest quality and least useful of the data types identified. 

High Frequency

High frequency data updates every 2-10 minutes, and like the low frequency data, is also sourced from mobile apps. Unlike low frequency data, this data can be used to estimate trajectories to some extent, and there is an ability to gather rough estimates per road segment for average speeds, though these estimates are not precise. This type of data also allows for exact travel time from origin to destination only. Similar to low frequency data, the update latency is 6 hours and there is no ability to view congestion, incidents or blockage in real-time. 

These factors make completing corridor, arterial and in-depth origin-destination studies nearly impossible, and would require the data in conjunction with other data sources to complete them. While this type of location data is slightly more precise than the low frequency data mentioned above, there is still only limited use for it. 

Connected Car Data

Location data from connected cars has become increasingly accessible with the connected car technology spike in newly made vehicles and fleets. Equipped vehicles can send their location and other data to the manufacturer or the fleet center, and this in turn serves us with expansive data for multiple uses and functions, including rich location data. This data is higher quality than the data coming from location-based services apps and can also be divided by frequency: 

Low Frequency

Low frequency connected car data updates every 30-90 seconds, which is much faster than the data from any location-based services apps. It is sourced from connected cars location-based services mobile apps, and fleet software. This data cannot determine exact trajectories, but they can be estimated. Users can complete full, precise origin-destination analysis with this data, though partial trip analysis requires some estimation. 

The major difference between this data and other types is the update latency, which stands at around 24-72 hours. This makes it extremely difficult to get the data relative to real time, and may impact any studies being completed with short turn around periods. Low frequency data can recognize congestion, incidents and blockage; however, this can only happen for historic data and per road segment.  

This data is much better quality than the previous types, but it does still have its restrictions. Users now have the ability to complete certain studies, like origin-destination, but many more complex analyses are still not accessible with just this data.

High Frequency

High frequency connected car data updates every 1-5 seconds and is sourced from connected cars, public transit, fleet, fleet software, and transportation network companies. Companies like Wejo can access data from over 10 million cars on the road, making it extremely useful for agencies trying to source meaningful metrics and analyses. 

Unlike the other types of data, high frequency data can determine exact trajectories with precision due to the location updating so rapidly. Users can see elements like speed changes and lane changes due to the precision of the location within the segments. It can also complete origin-destination analysis for full and partial trips, providing more in-depth location data. 

Exact speeds are extremely precise with high frequency connected car data, and users have the ability to find exact travel times within individual road segments. The update latency is every 15-60 seconds, so the data is actionable within minutes. Users can also see congestion, incidents and blockage in near real-time within road segments. 

High frequency data is not only higher quality and more reliable, but it provides the precision, speed, and accuracy required for active traffic management and road safety applications. This is the best and most complete data, which makes it the most useful for completing corridor analysis, signal performance measures, arterial analysis and other complex studies within particular segments. 

Having high frequency connected car data makes completing necessary analyses more accurate and provides better quality data, so that the decisions made can be relevant and timely. This data, available through SMATS, provides the most user flexibility and an accurate picture of how traffic is moving, how fast, to where, and for how long.  

Learn more about connected car data and how it is translated into ready-to-use visuals and insights in SMATS’ iNode