EDGE AI · AUDIENCE SIGNALS · PRIVACY

Proximity Targeting Using Demographics and Psychological Types

Study real-time demographic, gaze and engagement signals for content selection while treating consent, privacy and non-identification as system requirements.

Funded · Fiscal year 2026 · One-year project

THE RESEARCH STORY

Begin with the decision—not the model.

Digital media often relies on clicks or historical profiles, while physical spaces require decisions from transient signals. This project studies a microservice prototype that processes camera input, estimates audience-level signals and selects contextual content without requiring persistent personal identity.

RESEARCH INSIGHT

Context-aware communication may reduce irrelevant content, yet opaque inference about people creates its own risk. The project is therefore valuable only if it tests both effectiveness and the boundary it must not cross.

PRIMARY RESEARCH QUESTION

What must the study resolve?

How can a real-time analytics system create measurable business value while minimizing data collection, limiting purpose and preventing inference from becoming persistent personal surveillance?

OBJECTIVES

  1. 01

    Study real-time audience signals without creating persistent identity

  2. 02

    Develop a microservice architecture for inference, content selection and analytics

  3. 03

    Evaluate performance, fairness, privacy and appropriate operational limits

INTERACTIVE RESEARCH LAB

Explore the audience dashboard

Reconstructed student-project charts with simulated data. No live camera analytics or real billing.

METHOD ARCHITECTURE

From input to accountable support

The method is organized as an evidence chain. Every transformation should preserve enough context to explain what entered the system, what the model inferred, how confidence was assessed and when a person must review the case.

  1. 01

    Capture under appropriate notice and consent

  2. 02

    Person detection without persistent identity

  3. 03

    Audience-level and gaze-signal estimation

  4. 04

    Content selection and ranking

  5. 05

    Minimal aggregate logging and load testing

WORK PACKAGES

Four connected research responsibilities

Models

Study age-range, gender, gaze and engagement estimation with explicit bias and uncertainty reporting.

Architecture

Use microservices, Docker Swarm and RabbitMQ to separate services and support scaling.

Evaluation

Test latency, throughput, load, relevance and subgroup outcomes.

Governance

Design data minimization, retention, access control, notice/consent and stop-use procedures for elevated risk.

EVIDENCE PLAN

What the project should produce

A research project is stronger when its outputs can be inspected independently of the final model score. The dossier therefore separates artifacts from evaluation.

Expected research assets

  • Edge/microservice prototype for audience-level signal processing
  • Contextual content-selection module with reviewable rationale
  • Aggregate analytics dashboard with minimized personal data
  • Load, latency, fairness and failure-mode evaluation
  • Governance framework for notice, consent, retention and access

Evaluation dimensions

  • Latency and throughput
  • Signal-estimation uncertainty
  • Engagement lift under controlled tests
  • False inference and subgroup disparity
  • Data retained per interaction and deletion compliance

RESEARCH TRANSLATION

From prototype to a credible working system

Possible study contexts include digital signage, exhibitions, museums, service spaces and learning environments. Facial identification, high-impact decisions and inferences people cannot reasonably expect should remain outside scope.

01Laboratory evidence
02Relevant-environment validation
03Human and governance review
04Controlled operational pilot

LIMITATIONS AND RESPONSIBLE FRAMING

The boundary is part of the research design.

Demographic and psychological inference carries risks of error, stereotyping and discrimination. This presentation therefore frames the work around audience-level analytics, excludes facial identification, and rejects use for high-impact decisions.

This dossier communicates the documented project design and intended evaluation. It does not convert a proposal, funded status or prototype into a claim of validated operational performance. Numerical results should be cited only from a resulting peer-reviewed publication or verified project report.

PROJECT RECORD AND CITATION

A stable record for this research direction

FORMAL PROJECT TITLE

Proximity Targeting Using Demographics and Psychological Types

Suggested citation for this public dossier:

Arayaphan, W. (2026). Proximity Targeting Using Demographics and Psychological Types. Research project dossier, Warin Arayaphan.Use the original project report or resulting publication when citing methods, data or findings.

STUDENT PATHWAYS

Where students can enter the research

Student directions include edge inference, gaze estimation, distributed systems, load testing, fairness evaluation, privacy engineering and human-computer interaction.

Bring a research question.

We can shape it into a dataset, experiment, model, system component or responsible deployment study.

Talk to the lab →