Tartan CargoSmart Integration
Tartan CargoSmart Integration
Tartan uses data aggregation and analytics on AWS to provide Port-To-Port Container Visibility worldwide.
Tartan is a US company. Users of the service are also mainly in the US and internationally. The tracking of the container data and providing ETA is worldwide.
Tartan aggregates ocean carrier data and provides accurate, easy-to-consume container tracking services to carriers and their customers.
ETA of cross-ocean cargoes was off by several days without any options for the customers to make inquiries and improve on the ETA. Existing data streams were not reliable enough to provide more accurate information.
With tens of thousands of tracked containers per day and as many customers eager to get reliable ETA information on their shipments, Tartan needed a new data source integration to satisfy the customer demand.
SnapSoft has developed and provided the following solution on AWS:
- Deployed ECS service to pull data from the external provider, with autoscaling to handle
the arbitrary load
- ECS (Elastics Container Service) is a scalable container runtime. We
package the application services as separate docker images and use
ECS to autoscale our deployments based on the load, which is
proportional to the number of in-transit containers tracked. We set up
multiple services which scale independently:
- Data fetching services, which pull data from external sources, parse them and store them for further analysis. Cargosmart is one of these.
- Data analyzer service, which uses the loaded data, calculates estimated time of arrivals (ETAs).
- Front-end services that deal with direct user interactions on TartanHub, like logging into the system and viewing container details.
- ECS (Elastics Container Service) is a scalable container runtime. We package the application services as separate docker images and use ECS to autoscale our deployments based on the load, which is proportional to the number of in-transit containers tracked. We set up multiple services which scale independently:
- DynamoDB to store operational data
- DynamoDB is a managed serverless database offering.
- DynamoDB can automatically scale to any load size, so ideal for uneven / non-predictable usage, such as the Tartan container tracking service.
- Cloudformation for infrastructure
- S3 for file storage
- Cognito for user management
Results and Benefits
- Successfully integrated cargosmart
- Increased number of tracked carriers from 4 to 20
- Cargosmart is a service provider collecting shipping container data (GPS coordinates, events, harbors, ship names for a given container) directly from the shipping companies. They provide this data through an API which we integrated into the Tartan system to be able to pull container data for over 20 carriers.
- Tartan’s algorithm uses the historical data of previously completed shipments to estimate the time of arrival of an ongoing shipment. As the number of completed shipments increases, so does the amount of historical data to feed the algorithm and increase ETA forecast accuracy further.