RESEARCH PROJECTS

Applied AI begins with a decision worth improving.

This page presents funded or formally approved research projects only. Three projects apply artificial intelligence and data systems to operational decisions while keeping evidence, limitations, privacy and expert judgement visible.

01See the real problem

A model is one component inside a longer operational decision.

02Make evidence visible

Performance, uncertainty, provenance and failure modes belong in the design.

03Design for responsible use

Human review, privacy and appropriate boundaries are system requirements.

RESEARCH OPERATING MODEL

From a difficult question to evidence that can travel.

We treat publication, prototype and operational learning as connected outputs. The route is iterative: field evidence can revise the question, the data and the model.

01Observe the workflow

02Model the evidence

03Validate limits

04Transfer responsibly

FUNDED & APPROVED PROJECT PORTFOLIO

Three approved contexts. Evidence designed for use.

Each dossier documents the funded research question, architecture, work packages, evaluation, safeguards and ways students can make a meaningful contribution.

01DEEP CNN · AUTOMOTIVE DAMAGECar Damage Estimation System using Deep Convolutional Neural NetworkDetect visible damage, estimate preliminary severity and structure evidence for faster expert review.Funded · FY2023 02HEALTH AI · MORPHOLOGICAL FEATURESDevelopment of an Advanced Artificial Intelligence Model using Morphological Features to Enhance Cervical Cancer Screening EfficiencyCombine cell imagery, nucleus/cytoplasm segmentation and interpretable morphological evidence in screening research.Approved · Awaiting budget authorization 03PROXIMITY TARGETING · DEMOGRAPHICSProximity Targeting Using Demographics and Psychological TypesStudy real-time audience signals and scalable services while treating consent and privacy as architectural constraints.Funded · FY2026

Only projects documented as funded or formally approved are listed here. “Awaiting budget authorization” means the project has not started and does not imply that funds have been disbursed. Independent studies, prototypes and research archives remain elsewhere on the site.

FOR STUDENTS AND COLLABORATORS

Join through a question, not only a tool.

Contributions may begin with dataset design, annotation quality, evaluation, explainability, integration, privacy engineering or field validation. A focused contribution can become a thesis direction, a reproducible study or part of a working prototype.

Talk to the lab