A Practical Framework for Responsible AIoT (Artificial Intelligence of Things) Product Development in Impact Tech Startups

Abstract

Impact Tech Startups (ITSs) developing Artificial Intelligence of Things (AIoT) products operate in complex environments where technical feasibility, environmental sustainability, accessibility, and startup viability must be navigated simultaneously, often under conditions of uncertainty and with limited resources. Despite growing interest in the complexities of ITSs, existing guidelines for responsible product development remain overly broad and insufficiently adapted to be actionable for ITSs in early-stage product development. This study developed an initial framework integrating five core dimensions: Sustainable Development Goals (SDGs) alignment, accessibility and inclusion, environmental sustainability, AIoT technical considerations, and startup viability mapped across six stages of the Product Development Lifecycle (PDL). Each dimension is operationalized as decision-support questions structured through the four Responsible Innovation (RI) principles: anticipation, reflexivity, inclusiveness, and responsiveness. The study followed a Design Science Research (DSR) methodology comprising a narrative literature review, expert evaluation, and an exploratory application with DiAmíti, a Costa Rica-based ITS. The findings confirmed the need to adapt existing guidelines into practical tools aligned with the realities of product development in these contexts. The framework received relevance ratings of 4.5/5 from experts and 4.3/5 from DiAmíti. The primary contribution is a novel first-iteration unified framework that integrates these five dimensions within AIoT product development. Future work should extend the framework through additional iteration cycles based on the insights gathered from the evaluations and broader testing across diverse ITS contexts.