Smell the future

Episode Summary

Executive Summary: This episode of BBC Tech Life explores three ways technology is revealing hidden worlds: AI-powered odor sensing that could detect disease, sleep-tracking devices that promise deeper health insights but have limits, and bird ID apps that turn casual nature walks into data-rich citizen science. Across segments, the show highlights promise, accuracy tradeoffs, and ethical questions about privacy and environmental monitoring.

Main Topics: AI and electronic noses for odor detection (Priority: 5/5): Professor James Logan explains how DOT uses sensor and AI systems to decode complex smell mixtures into machine-readable data, aiming to give devices a sense of smell for health, agriculture, and pest detection. Disease detection through odors (Priority: 5/5): The program examines how smell-based sensing could identify illnesses such as COVID, malaria, and dengue, and potentially detect cancer, drawing on both human studies and animal sniffing research. Conservation tech and puffin monitoring (Priority: 4/5): DOT is working with the Scottish Government to detect puffins, chicks, predators, and bird flu in burrows using odor sensors, reducing the need for invasive fieldwork on remote islands. Privacy and ethics of ambient sensing (Priority: 5/5): The segment raises questions about who owns odor data emitted by people and how to protect sensitive biological information when sensors in homes, buildings, or wearables constantly sample the air. Sleep-tracking wearables and under-mattress sensors (Priority: 4/5): Shona McCallum reviews consumer sleep tech, including rings, watches, and Withings' under-mattress sleep analyzer, which estimates sleep patterns and may help spot issues like sleep apnea. Bird identification apps and younger users (Priority: 4/5): Liv McMahon shows how Merlin Bird ID and similar apps use sound recognition and crowdsourcing to make birdwatching more accessible, especially for younger audiences, and to support conservation data collection.

Key Arguments: Smells contain rich biological information that humans usually miss, but AI and sensor platforms can decode odor signatures by analyzing chemical ratios rather than single compounds. Electronic noses could become practical tools for early health warnings in wearables, homes, farms, and hospitals, with applications ranging from disease screening to pest detection. Animal and insect sniffing research has helped inspire machine-smell systems; dog-based detection work, including COVID sniffing, validates the concept. Puffin conservation can benefit from odor sensors because current monitoring methods are dangerous, labor-intensive, and invasive on remote islands. Ambient odor sensing creates serious privacy concerns because biological information is constantly released into the air, unlike blood samples that people knowingly provide. Sleep trackers are popular and can track long-term trends, but they do not match clinical polysomnography for accurately measuring sleep stages like deep sleep and REM. Bird ID apps lower barriers to wildlife observation, helping younger users get involved and generating large-scale conservation data through crowd science.

Data Points: Human body volatile chemicals: around 600-700 - Estimated number of volatile chemicals emitted from the human body through sweat, skin cells, and skin bacteria. Sleep tracking among Americans: nearly half - Reported share of Americans tracking their sleep. Sleep tracking in the UK: 48% - Reported share of people in the UK tracking their sleep. Sleep apnea concern in UK: 1 in 5 people - Withings cited this as the proportion of people in the UK concerned with or showing signs of sleep apnea. Men with signs of sleep apnea: 26% - Withings cited this as the estimated proportion of men affected. Women with signs of sleep apnea: 14% - Withings cited this as the estimated proportion of women affected. Merlin downloads this year: at least 12 million - App usage figure mentioned by Liv McMahon for Merlin Bird ID on smartphones. Bird sightings recorded by Merlin: 4.4 billion - Total sightings data accumulated by the Merlin Bird ID app. DOT odor detection examples: COVID, malaria, dengue - Diseases Professor Logan said DOT sensors can detect with high accuracy. Puffin monitoring method: meters into the ground - Traditional conservation method described for checking puffin burrows on remote islands.

Pivotal Quotes: "we actually live in a smellscape" — Professor James Logan: Explaining DOT's view that odor is a data-rich environment that machines can learn to interpret. "Shazam for birds" — Jessie Barry: Describing Merlin Bird ID’s sound identification feature that listens to bird calls and identifies species nearby. "There is a big ethical question around how do we, you know, if we have sensors and Buildings, for example, that are sampling from people. How do we actually ensure that the data is safe" — Professor James Logan: Discussing privacy risks of constant odor sensing and ownership of biological data in the air.

Implications: The episode suggests sensing tech will increasingly move from niche novelty to everyday health and environmental monitoring. But it also signals that accuracy, long-term validation, and data privacy will be central challenges as these tools scale.