Description of original award (Fiscal Year 2022, $600,984)
In this new NIJ solicitation, our goal is to further develop an existing targeted mobile forensics platform developed under Grant Number NIJ-2016-8976. The system called Targeted Data Extraction System (TDES) was built for extracting case specific data in a forensically sound manner from a mobile device during triage, from willing participants. Here we plan to build a new ``forensics intelligence" platform by extending this system by augmenting it with substantial new capabilities. In this project we set a new goal of significant development in building a more functional, robust, and practically useful selective data extraction system for law enforcement, with the goal of providing forensic intelligence during events of mass incidents, like the events during the Boston Marathon bombing or the Aurora mass shooting. At the same time, we propose a moderate applied research goal to develop new and interesting features of the system that will make it much more effective and practical through the addition of cutting-edge AI capabilities. More precisely, the new project will focus on building a practical and scalable multiphone data extraction system with forensic intelligence capabilities, which we call SM-TDES. TDES has already been going through an evaluation by NIJ's Criminal Justice Testing and Evaluation Consortium (CJTEC), led by RTI International. The primary technical interface for this evaluation has been Mr. Robert O'Leary – a well-known digital forensics subject matter expert. With respect to the development goal of SM-TDES, we would hope to be able to complete any suggestions for improvements that arise during the evaluation process while at the same time augmenting the TDES system significantly. Our proposed SM-TDES under this new solicitation would focus on three criteria: (1) we want to highlight and explore the need to extract data from many phones concurrently for any type of mass incident involving many people; (2) we want to develop significant analysis capabilities using AI to automatically extract information of relevance for pursuing leads in mass incident cases; it is likely that even the selectively extracted data from multiple phones is expected to be of large volume; and (3) we want to ensure that both the smartphone interface and the back-end analysis system are simple yet robust and functional to the point that it is possible to use the resulting system in real environments.