Here Prototype Solution

Our main idea was to create a data validation system that detects and fags errors in POI datasets using three key filters: Existence Verification - Checks if the POI still exists (based on recent activity or source data) Geolocation Accuracy - Confirms if the POI is placed on the correct side of the road or within a valid area. Identify and verify correct side POI location We implemented a system that cross-references raw POI data with: Official databases Updated digital maps Logical rules We used a combination of: Python Pandas to clean and process the datasets. A step-by-step pipeline: Detect empty or null fields Validate POI coordinates within allowed boundaries Match POIs to their closest valid road segment (based on direction). GitHub or collaboration and version control To handle POIs that no longer exist, we implemented a validation filter that checks for missing or outdated information, such as empty fields, invalid coordinates, or lack of recent verification. If POI fails all validation criteria it’s lagged as obsolete and marked for deletion As a result of the previous step , we now have less information to handle with. To approach the first problematic our idea was to find and filter duplicated values and identify if one has been moved since in the last step we already deleted the non-existing ones. For the second scenario where the POI it´s in the wrong location. With the node of reference to know wich is the correct side The random distance is uncertain so we rule out that proposal For the third scenario we think about compare the direction travel and the multiply digitised Actually we realized that we were going down the wrong path, maybe we needed to have this map to help us visualize the data graphically and interpret it better. In the end we started this map, but the processing of the data was very slow due to the large amount of information, so it was not sustainable.