WASHINGTON - The Department of Homeland Security (DHS) Transportation Security Laboratory (TSL) is seeking industry input on a cloud-based data-sharing platform designed to support the development and validation of artificial intelligence-enabled threat detection algorithms for transportation security screening systems.
Through a request for information (RFI), DHS Science and Technology Directorate’s Office of National Laboratories Transportation Security Laboratory is exploring solutions to establish a Screening System Data Sharing Consortium (SSDSC) that would facilitate the collection and sharing of screening system data ranging from non-sensitive information to Sensitive Security Information (SSI).
The effort supports the Transportation Security Administration’s (TSA) transition to an open-architecture approach for transportation security equipment (TSE), emphasizing standards-based, interoperable hardware and software components. According to the RFI, this approach is intended to enable a broader range of industry partners to develop innovative subcomponents and support adoption of new screening technologies.
Related: What it will take to certify AI for commercial aviation
TSL conducts research, development, validation, test and evaluation, assessment, and certification testing of transportation security equipment designed to detect and mitigate threats involving explosive devices, contraband, and other materials. Through the SSDSC effort, TSL aims to provide authorized developers with secure access to well-curated datasets needed to train and test machine-learning-enabled threat detection algorithms.
Cloud-based tech
The proposed cloud-based platform would serve as a centralized repository for data collected from developmental screening systems and deployed transportation security equipment. Data sources would include X-ray computed tomography (CT) explosive detection systems (EDS) used for checked baggage and carry-on baggage screening, as well as millimeter-wave passenger screening systems.
The repository also would include augmented and fully synthetic datasets, which can help developers train and evaluate artificial intelligence models while addressing challenges associated with obtaining sufficient quantities of representative real-world data.
Beyond data storage, the envisioned platform would provide a controlled environment for developing, testing, and validating detection algorithms. DHS is seeking industry input on whether algorithm developers would receive access to a cloud-based testing environment using containerized code running on cloud-based central processing units (CPUs) and graphics processing units (GPUs), or whether developers would download datasets for testing on their own computing platforms.
Related: How AI is already changing commercial aviation—and why it still isn't flying the airplane
The RFI identifies several technical objectives for the potential platform, including secure migration of screening system datasets from existing local network-attached storage systems into a cloud environment while maintaining data integrity. The system would need capabilities for data ingestion, verification, annotation, redaction, permission management, and version control, along with audit trails documenting data access, annotation, and modification.
Facilitating collaboration
The platform also would support collaboration among SSDSC members and other approved stakeholders through secure application programming interfaces (APIs) for data ingestion, curation, annotation, modification, and access. DHS is seeking information on methods for allowing developers to access datasets for training and analysis, including potential connections to developers’ private cloud environments.
Because the data environment could include information ranging from publicly available datasets to SSI, DHS is seeking approaches for maintaining access controls, tracking user activity, protecting data privacy, and ensuring compliance with applicable security requirements.
The agency also is requesting information on approaches for managing costs associated with platform operations, including storage, downloading, and cloud computing services. Potential solutions could include cost allocation, cost-sharing, or chargeback mechanisms among government and consortium participants.
The proposed SSDSC reflects a broader aerospace and defense technology trend in which artificial intelligence development increasingly depends on access to large, representative datasets and secure computing environments. Similar challenges exist in autonomous systems, intelligence and surveillance applications, and other sensor-driven platforms where algorithm performance depends on continuous training, validation, and improvement.
For aviation security systems, an open architecture approach could allow screening equipment providers, software developers, and research organizations to develop and validate new detection capabilities without requiring complete replacement of deployed hardware.
DHS requests responses to this RFI by 18 August 2026 at 3:00 p.m. Eastern. The agency named Robert Marosz as the primary point of contact for this inquiry. They can be reached via email at [email protected]. More information is available at https://sam.gov/workspace/contract/opp/052ce68548d945098b4fb8fd7f992918/view.
