iVolve sites across the world are home to databases containing billions of GNSS locations from their real-time fleets.
Our systems actively monitor machine location, and fuse this data to a wide range of other machine information to provide real-time insights. In this way, an immense amount of value is already extracted in real-time use cases of location data. But what’s interesting is that after these initial insights have been extracted, the wealth of position data is generally forgotten.
We now find ourselves dealing with a mountain of historical location data, a quintessential big-data problem. Our challenge is not how to process or store this location data, but rather how to reimagine it, and look deeper for new insights.
Let’s consider one potential use here, and focus on mine site road maintenance. What might we be able to glean about roads from a database of past location records?
Our first step would be to clean up the data. Inaccuracies are inherent in LP-GNSS systems, so we begin by purging our dataset of all outliers. Statistical analysis methods, such as density-based clustering, could be utilised to remove those trajectories that have been distorted by traditional GNSS errors. We might then pull in tools like Bayes classifiers to extract lane and intersection information from the location data.
The end result of some clever data processing would be a 3D centimetre-level accurate model of your road network on site!
Things only get more interesting from here when we consider the data that we can then overlay onto the model. Imagine road use quantified and visualised in terms of the number and type of vehicles, time of day, vehicle speeds, road gradient, etc! We could then augment the model further by pulling in other datasets from the iVolve Database. Overlaying information like road condition events, strut pressure spikes, tyre pressure, and more!
What’s more this real-world road model could be generated for any span of time encompassed by the iVolve database, allowing comparisons of past and present roadways.
And that’s just one example of an insight waiting to be unlocked at an iVolve enabled site.
To stay up to date, follow Callum on Twitter here.