AI-Driven User Experience and Accessibility Enhancements for a Sensor-Based Platform
Loading...
Authors
Sharma, Nikhil Nick
Issue Date
2025
Type
Thesis
Language
en_US
Keywords
Accessibility , Aritificial Intelligence , Environmental Monitoring , IoT , Sensors , User Experience
Alternative Title
Abstract
The complexity and volume of IoT-generated data have grown rapidly, outpacing the ability of non-expert users to interpret and act on it to make real-world decisions. This thesis presents a prototype platform that integrates artificial intelligenceto enhance both the user experience and accessibility in data-intensive environments. The system implements a conversational large language model capable of generating real-time, natural language summaries of various types of data visualizations-making complex information more accessible to users, particularly those relying on screen readers or keyboard-only interactions. A comprehensive assessment of the system and data interpretation capabilities demonstrates its reliability and preferred use by users. Beyond data summarization, the portal enables further querying of data, allowing users to converse with AI using natural language to explore insights without needing expert knowledge on the domain. The interface possesses full keyboard accessibility design and a suite of guided help videos to accommodate a wide range of accessibilityneeds. Additionally, a map-based visualization layer was launched, allowing users to explore live sensor data and receive AI-driven insight related to the geographical positions of the sensor. The prototype has been applied to a platform for environmental monitoring, but demonstrates potential across domains such as agriculture, healthcare, financial management, and citizen science.