AI Wearables Face Major Adoption Hurdles Despite Tech Advances
Review reveals 30% of seniors abandon wearables within weeks, highlighting usability gaps in smart healthcare devices.
Summary
This comprehensive review examines AI-enabled wearables and home diagnostics through the Pi-CON methodology framework. Despite a forecasted $39 billion market by 2026, adoption challenges persist. Over 30% of seniors discontinue wearable use within two weeks due to setup difficulties and discomfort. The review analyzes vital sign monitors, digital diagnostics, and body composition tools, finding accuracy varies widely based on conditions and populations. AI plays a crucial role in passive monitoring systems like camera-based photoplethysmography and radar vitals detection. The Pi-CON framework emphasizes passive, non-contact, continuous monitoring to reduce user burden and improve long-term engagement.
Detailed Summary
The global AI-powered wearables market is projected to exceed $39 billion by 2026, driven by an aging population where 20% of Americans will be over 65 by 2030. However, this technological promise faces significant real-world adoption barriers that threaten to limit clinical impact.
This narrative review synthesized literature from 2020-2025 using the Pi-CON methodology framework, which evaluates passive, non-contact, and continuous monitoring systems. The authors analyzed three key categories: vital sign monitoring wearables, digital diagnostics, and body composition assessment tools, examining their technical performance and user experience challenges.
Key findings reveal substantial usability gaps despite technological advances. A 2024 study found that over 30% of senior participants failed to meet usage expectations during a two-week wearable trial, reporting difficulties with device setup and physical discomfort. Accuracy varies significantly across devices and conditions - heart rate measurements are generally reliable at rest, but SpO2 and respiratory rate show high variability. Optical sensors particularly struggle with motion artifacts and skin tone variations.
The review highlights promising developments in passive monitoring technologies. Camera-based photoplethysmography, radar-based vitals monitoring, and smartphone body composition apps like Spren (showing r≈0.96 concordance with DEXA scans) demonstrate potential for reducing user burden. AI algorithms enable real-time artifact filtering, anomaly detection, and personalized insights without manual data input.
Clinical implications center on the gap between technological capability and real-world implementation. While devices exist to monitor multiple vital signs continuously, their effectiveness is limited by human factors rather than hardware limitations. The authors recommend focusing on passive, unobtrusive systems that integrate seamlessly into daily life, particularly for aging populations and chronic disease management. Success requires addressing usability barriers, ensuring diverse training datasets to avoid algorithmic bias, and maintaining regulatory oversight as these tools transition from wellness to diagnostic applications.
Key Findings
- Over 30% of senior participants discontinued wearable use within a two-week trial period due to setup difficulties and discomfort
- AI-powered wearables market forecasted to exceed $39 billion by 2026, with 44.5% of US adults reporting planned wearable use
- Spren smartphone body composition app achieved r≈0.96 concordance with DEXA scans and ~2.3% mean absolute error across 5,500+ users
- Pi-CON-based non-contact sensors showed 0.33 vs 0.85 operator errors per measurement compared to conventional patient-generated health data devices
- Heart rate measurements generally accurate at rest, but SpO2 and respiratory rate show high variability across consumer devices
- 20% of US population expected to be over 65 by 2030, driving demand for home-based monitoring solutions
- AI model bias demonstrated when white individuals were overrepresented in training datasets, resulting in poorer accuracy for Black participants
Methodology
This narrative review followed SANRA criteria and synthesized literature from PubMed, IEEE Xplore, ScienceDirect, and Google Scholar for studies published 2020-2025. Search terms included 'AI in healthcare,' 'wearable diagnostics,' 'non-contact sensors,' and 'user engagement in digital health.' The Pi-CON methodology framework was applied to evaluate passive, non-contact, and continuous monitoring systems. Studies were selected based on relevance, quality, and originality, with older foundational references included selectively.
Study Limitations
This narrative review acknowledges that accuracy and usability findings vary significantly across device types, populations, and use conditions, making generalizations challenging. The Pi-CON methodology, while useful as a framework, represents one approach among many emerging multimodal sensing paradigms. The review notes that regulatory oversight varies widely, with some devices requiring FDA clearance while others remain in wellness categories with limited clinical validation. No external funding was received for this research.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.