Multimodal Objective Sensing to Assess Individuals with Context (MOSAIC)

Selecting and evaluating a workforce that is well-suited for the psychological and cognitive demands of the diverse positions across the Intelligence Community (IC) is an important and persistent need. This is growing in importance as the pace and complexity of the challenges facing the IC workforce grow and expand. Methods that enhance our ability to evaluate an individual’s psychological drivers, cognitive abilities, and mental wellness and resilience will enable improved capabilities to select the right person for the right job, evaluate and help maintain optimal performance throughout their career, and better understand and anticipate changes in an individual that may impact their work effectiveness, productivity, and overall health and wellness.

To address this challenge, the MOSAIC program aims to take advantage of multimodal mobile, worn, and carried sensors and the corresponding data to enable the measurement of an individual in situ, throughout their daily activities, using an aggregate of behavior, physiology, social dynamics, physical location and proximity, as well as other novel data sources. Research in this program will aim to establish convergent validity of multimodal signals across a range of researcher-defined contexts and over time to enable accurate and personalized evaluations. It is anticipated that research teams will develop and test a suite of multimodal sensors to collect a range of subject-focused and situational data; build capabilities to develop an integrated model of the subject, their behaviors, and the social and physical context; and advance methods to personalize modeling approaches to develop accurate assessments of an individual over time.

MOSAIC is anticipated to be a two-phase program. Briefly, each phase will focus on the following, with progressively more challenging metrics:

  • Implementation and testing of a suite of mobile, worn and/or carried sensors to measure the target constructs using data collected from diverse volunteer participants that provide informed consent to measure them during daily activities, along with more controlled data collection periods.
  • Development of methods to measure a subject, their social environment, and physical surroundings to inform the interpretation of a subject’s own data.
  • Advancement of methods to build automated and personalized assessments of the target constructs for each subject.

In both phases performance against target metrics will be evaluated by an independent test and evaluation team.

Note that all research will require approval from the appropriate Institutional Review Board(s), informed consent from subjects, and offerors will be asked to propose and implement the necessary safeguards to protect and maintain subjects’ privacy, data, and other legal, ethical, and safety protections.

Collaborative efforts and teaming among potential performers will be encouraged. It is anticipated that teams will be multidisciplinary, including but not limited to expertise in behavioral, cognitive, and psychological science; mobile computing and pervasive sensing; signal processing; data science; machine learning; data privacy and security. Work that will not be funded under the MOSAIC program includes: the development of new devices or sensors; improvements in device or sensor size or weight; advances in power sources or components; advances in radio technology; advances in materials; and biometrics.

IARPA anticipates that academic institutions and companies from around the world will participate in this program. Researchers will be encouraged to publish their findings in academic journals but all publications will require a pre-publication security review.

Contracting Office Address

Office of the Director of National Intelligence
Intelligence Advanced Research Projects Activity
Washington, District of Columbia 20511
United States

Primary Point of Contact:
Alexis Jeannotte
Program Manager