This is a Data-oriented project with the objective to provide a powerful tool for getting insights from traffic data in Antofagasta using the Waze API through its program cities, which provides currently active Alerts and Jams.
Page is available in https://traficoantofagasta.com
The data is provided in JSON format from the API and contains Alerts and Jams events. Alerts represent all events reported by users and Jams are traffic congestion events that Waze calculates from average speed, quantity of vehicles, etc. This application is centered on Alerts events because they include both jams and user-reported alerts, and additionally provide more information, such as ACCIDENT or ROAD_CLOSED types. The complete information is available in Waze documentation.
- Obtain sufficient and representative data about traffic events in Antofagasta through the Waze Cities API.
- Perform a descriptive analysis of the collected data to identify relevant patterns and trends in traffic behavior.
- Identify key factors that influence road safety and traffic efficiency in the city.
- Present visually comprehensible and useful information for traffic management authorities, facilitating the implementation of data-driven policies and actions.
- Use available data to develop predictive traffic models that would allow anticipating events and making proactive decisions in traffic management.
The app is composed by:
- Local server
- Client (Dashboard)
The local server, written in Rust, is in charge of all core processes related to data extraction, transformation, and loading to local database (ETL), including cache management (Memcached). The server provides endpoints for interacting with the client (Dashboard).
http://localhost:7000/update-data
http://localhost:7000/get-data?since={millis_epoch_time}&until={millis_epoch_time}
http://localhost:7000/clear-cache?key={key}
http://localhost:7000/aggregate
update-data:GET- endpoint retrieves data from the API, transforms it, saves to database and cache. Then returns the retrieved and transformed data in theHTTPresponse.get-data:GET- endpoint that provides data from cache if available or from the database, with parameterssinceanduntilwhich indicate the period ofpub_millis(publication date) in epoch time and milliseconds.clear-cache:GET- endpoint that removes data from a key in cache.aggregate:POST- endpoint for transforming data from native API form to aggregated and transformed data. The request body should contain the alert data in aJSONwith the keyalertscontaining a list of all the correctalertsdata.
The client (dashboard), written in Python, is in charge of retrieving data from the local server, deploying the dashboard, listening for callbacks and updates, and training/deploying the Machine Learning model.
flowchart TD
subgraph Waze
direction LR
A[Waze user] --> B[Alerts] --> C[Waze database] --> API[Waze Cities API]
end
subgraph App
direction LR
S[Server] --> D[Local database] --> E[Dashboard] --> F[Final user]
end
Waze --> App
classDef waze fill:#fff0,stroke:#333,stroke-width:0.5px,stroke-dasharray: 5,5;
class Waze waze;
class App waze;