ADELE

Artificial Intelligence-Driven Data Enhancement Platform with Natural Language Enabled

Cloud-hosted data enhancement and integration platform governed by business rules.

Machine learning capability to continuously increase and optimize processing logic and efficiency.

Data conflation program that achieves 97% automatic match rate and greatly reduce human labor.

Catches and automatically fixes most errors, yielding 99.99% accuracy, much higher than industry standard.

Automatic geospatial feature realignment software that enables user to realign roads to centerline of roads based on orthophotography with 99% accuracy.

Easily integrated with Internet of Things (IoT) through extensible APIs.

North American Rail Network Data Consolidation

By using ADELE, SFI’s IT professionals and data consolidation experts helped Federal Railway Administration (FRA) consolidate the entirety of North America's rail network data. From multiple sources with different owners, purposes, geospatial accuracies, and non-geospatial attributes, to a single comprehensive, standardized, and high-resolution dataset, our innovative staff work closely with U.S. Department of Transportation, Department of Defense, and other stakeholders in building a better rail network dataset for the nation to use in years to come.


New Jersey National Hydrography Dataset Creation

By using ADELE, New Jersey’s National Hydrography Dataset (NHD) stewards improved the data resolution 10 folds – from the previous 1:24,000-scale to 1:2,400 -scale local resolution hydrographic data delineated by aerial photo interpretation. Around 6,000 new miles of flowline and 3,000 new waterbody features were added to the dataset. The project was a tremendous success. New Jersey was the first state that has local resolution data in the multi-resolution repository.


Facility Data Enhancement for Utility Company

SFI helped a utility company (name confidential) realign its facility data in a “no downtime” production method. Link Vector Creation and Triangular-Irregular-Network (TIN) Realignment are the two modules in ADELE that were heavily used in this project.