[Inside Nomadia] - Nomadia Data Center: the data experts who make your applications possible

You use Nomadia solutions on a daily basis, but how much do you know about the data that is used to geocode your addresses, calculate your routes and optimize your routes? At the head of Nomadia’s Data Division, Xavier Roumanie explains where this data comes from, how it is integrated, updated and made available in your tools to meet your operational needs – without you having to deal with it. 

What is the role of Nomadia’s Data Department?

This is a completely cross-functional role since all Nomadia products are based on data sets without which our software would be comparable to fuelless engines. When you use a Nomadia solution, you’re really using two things: a software engine that provides a number of services and features, and reference data, i.e., user-agnostic, that feeds that engine.

The team I lead is made up of mapping engineers and developers who produce this ‘fuel’ from raw data from different suppliers. The core of our work is to transform this data and integrate it so that it can be directly used by the functionalities of our software, whether it is to display maps, locate addresses, calculate routes and plot isochrones or schedule routes.

What types of data are involved?

They can be classified into 3 categories. First of all, there is data with a geographical/cartographic dimension. They form the basis of applications by making it possible to locate and characterize roads, buildings and all elements that have a position in space. For this category of data, we work mainly with two providers: HERE, which is one of the world leaders in the location data market and whose raw data we buy; and, on the national perimeter, the IGN, whose data is now freely accessible. Data from the OpenStreetMap collaborative database can be added to this. In addition to global coverage and hundreds of categories of points of interest, HERE has the advantage of offering a wide range of attributes that characterize the road network in a very detailed way, which enhances the accuracy and reliability of downstream route and route calculations.

We also integrate what we call ‘attribute’ data, in particular those from INSEE, such as census data using the IRIS grid, with more than 1200 variables, and the SIRENE file which provides information on 11 million establishments classified by activity codes (NAF code). It is this data that allows users of our solutions to carry out catchment area studies or catchment area analyses based on the socio-demographic characteristics of households.

Finally, we integrate the address points of the National Address Base (BAN) which is the only address database officially recognized by the administration with today more than 25 million addresses located with the greatest precision.

How does this raw data become usable by Nomadia solutions?

For HERE and IGN data, we have developed an automated processing chain that takes raw data from our suppliers as input and puts it in Nomadia’s own format. At the end of the day, the Data Division delivers 3 types of products:

• “display” files that are used to display maps;
• “routing” files, i.e. graphs that contain all the data needed to calculate routes, isochrones, distances, etc.
• georeferencing or “geocoding” files which, from an address, make it possible to obtain its geographical coordinates (X,Y) and to position this point on the map.

As IGN and HERE publish updates of their data every quarter, new versions of these 3 products – display, routing and geocoding – are made every quarter in order to provide our end customers with the most up-to-date data. Before being distributed, these new versions are tested on all of our software to ensure their quality and avoid any risk of regression.

How do customers have access to this data?

Today, the majority of our customers use SaaS versions of Nomadia solutions. They automatically get quarterly updates, with no need to download or install anything. Access to and use of the data is done via what we call ‘GeoAPI’. These are web services that cover the 3 basic functions present in all our solutions: map display, routing and geocoding.

How are Nomadia graphs developed and what is specific about them?

The graph is the indispensable element for the calculation of routes and routes. In particular, the data quality of the graph determines the average speeds on each road section, depending on the type of lane and the type of vehicle. This is what allows our software to calculate realistic travel times. For maximum accuracy, we use many attributes from the HERE road database to determine “speed patterns” corresponding to the specifics of each section of track. By cross-referencing attributes such as the maximum speed allowed, the number of traffic lanes, the presence or absence of a lane separator, the direction of traffic, the urban/non-urban environment, we established a typology of about 350 road families.

Each of these road categories is associated with categories of vehicles, bearing in mind that the average speed of a car will be higher than that of a light commercial vehicle, which in turn will be higher than that of a heavy goods vehicle. In addition to these vehicle profiles, we apply traffic conditions according to days and times. We have established 5 levels that range from the smoothest traffic to traffic jams. Cross-referencing all these parameters and integrating all traffic restrictions/constraints into our graphs (limited height and width, one-way streets, cycle paths, pedestrian areas, etc.) results in travel times that are much closer to reality than if we only took into account the speed allowed, as is the case with some applications. This is fundamentally what enables our customers to establish reliable routes, whether they are deliveries or field interventions.

What have been the major developments in recent years in the field of geographic and road data?

The most striking phenomenon is the development of open data. More and more data is open and freely accessible. This is evidenced by the opening up of data from the IGN and INSEE. The BAN is another example. This address database was previously managed by the IGN. It is currently managed by the Interministerial Digital Directorate (DINUM) and fed directly by the municipalities, which can at any time integrate new roads, new numbers, new building addresses or changes in street names. Enriched daily, BAN offers an unmatched level of accuracy for all geocoding applications.

Collaborative data has gained and continues to gain ground. Open Street Map is the archetype, but HERE is no slouch. Its Map Creator tool allows a large community to create and edit speed, flooring, numbering, and more data. to increase the richness and accuracy of its databases.

The acquisition of data through vehicles is also a notable development. This is particularly the case at HERE, which has belonged to the consortium of German manufacturers BMW, Mercedes-Benz and Audi since 2015. This feature allows it to provide up-to-date data on the road network from tens of millions of vehicles around the world. HERE also collects increasingly accurate data such as the winding and gradient of road sections, which influence the calculations of travel times and the range of electric vehicles.

All these developments enrich our data models and allow our customers to make calculations that are ever closer to the reality faced by their employees on the road.

How did the Data Department take into account the issue of the Olympic and Paralympic Games?

The traffic disruptions related to the Games were very worrying for our customers. We have integrated the restrictions associated with the tests in the form of ‘reject flags’ in our graphs in all the areas concerned. This data is part of the Q2 update that will be available at the end of June. This update will take into account all possible scenarios, with restrictions and exemptions varying from day to day depending on the venue and the event calendar. All users will have access to this data, without restriction. To calculate their routes during this entire period, they will just have to choose the right vehicle profile and tick the JO option in their software. A few days before the event, a warning will be sent to customers via bubble info to alert them to the various disruptions planned and facilitate the management of traffic conditions. The principle is the same as for EPZs, the only difference is that these special provisions are limited to the duration of the Games.

Do you have any other special requests?

Absolutely. Between 10% and 15% of the Data Division’s activity consists of providing services for customers with specific needs, whether recurring or not. These are tailor-made services, such as geocoding a million addresses at the most precise level or standardizing very poor quality address files. Some clients call on our expertise to determine which branch in their network makes the most sense to link each of their clients to. Others ask us to create their own sectorization. The demands are extremely varied. In addition, there are those of the Pre-Sales and Professional Services teams, for whom we assess and integrate data provided by the customer, which sometimes leads us to develop specific features and tools.

However, the core of our business remains everything I have described since the beginning of this interview: integrating data, updating it and making it available to all our customers so that they can organize their own business based on the latest and most reliable data at all times.