Request for Information (RFI) - Artificial Intelligence for Image Adjudication

Background
The U.S. Department of Homeland Security (DHS) Customs and Border Protection (CBP) is focused on enhancing its Non-Intrusive Inspection (NII) systems as part of its multi-layered enforcement strategy to detect illicit merchandise and substances without physically opening or unloading cargo, baggage, and conveyances.

This RFI aims to gather information on Artificial Intelligence (AI) and Machine Learning (ML) tools that can assist in the adjudication of images generated by large-scale imaging systems used for scanning privately owned vehicles (POVs) and cargo vehicles. The goal is to improve efficiency, security, and throughput in border security operations while maintaining the flow of legitimate trade and travel.

Work Details
CBP seeks AI/ML tools to assist in NII image adjudication, focusing on tasks such as anomaly detection using a human-in-the-loop approach. Key objectives include:

1. Implementing AI/ML models for computer-assisted analysis of NII images to enhance operational effectiveness, particularly for high volumes of LS-NII scans.
2. Ensuring that proposed solutions integrate with existing CBP data platforms rather than requiring new proprietary systems.
3. Automating the identification of anomalies, conveyance modifications, manifest validation, contraband identification, and commodity verification through algorithms applied to LS NII imagery and metadata.
4. Supporting officer image analysis on higher-risk shipments and identifying areas of interest for efficient operations.

Place of Performance
The contract work will be performed within the operational environments of CBP facilities across various geographic locations where NII systems are deployed.