It is often thought that the closer to your body a health tracker is, the more accurate the data. However, this assumption has been turned on its head by scientists at King’s College London, who have found a way to track health without compromising comfort. Their research shows that loose clothing can be used to capture body movement more accurately than tight body straps.
Based on patient feedback, the team realised that many motion-tracking devices are rigid, uncomfortable, and intrusive. This was especially problematic for patients during rehabilitation, where “ease of dressing and long-term wear really matter,” said Dr Matthew Howard, paper co-author of the research published in ‘Nature Communications‘ and a Reader in Engineering at King’s College London. Taking this feedback on board, the researchers set out to find an alternative approach that would be accurate, non-invasive, and easy to use.
“We wanted to see whether the natural movement of loose clothing, something we know people already wear, could be used as a useful signal rather than treated as noise. This addresses a major gap: creating motion sensing systems that feel genuinely ‘everyday’ rather than medical, increasing their adoption in the process,” said Dr Howard.
As the research progressed, the team discovered that loose fabric can predict and capture the body’s movements with 40% more accuracy while needing 80% less data than sensors attached directly to the skin. This finding ran counter to long-held assumptions in wearable technology.
Dr Howard said: “When we think about technology that tracks movement – like a Fitbit on your wrist or the suits actors wear to play CGI characters – we had thought that the sensors need to be tight against the body to produce the most accurate results. The common belief is that if a sensor is loose, the data will be ‘noisy’ or messy.
“However, our research has proven over multiple experiments that loose, flowing clothing actually makes motion tracking significantly more accurate. Meaning, we could move away from ‘wearable tech’ that feels like medical equipment and toward ‘smart clothing’ – like a simple button or pin on a dress – that tracks your health while you feel completely natural going about your day.”
According to Dr Howard, this approach is particularly well-suited to applications where comfort and long-term wear are essential, including fall detection and home-based rehabilitation. However, because the system relies on how fabric moves with the body, it is “less suited to environments with strong, unpredictable airflow that could distort the readings – like outdoor sports in high wind, such as paragliding.”
The challenge of clean data
Fabric does not move in a fixed or predictable way, which makes modelling its deformations challenging. To address this, the team approached the problem statistically.
“By simulating a range of movements in a controlled virtual environment, we built an understanding of the main repeating patterns of movement and trained algorithms to recognise them. This allowed us to separate useful motion cues from irrelevant cloth behaviour,” said Dr Howard.
An unexpected result
Sensors are often designed to sit tightly against the body to detect subtle physiological signals such as pulse, breathing, or blood flow. Yet the research showed that when a sensor is too tightly coupled to the body, micro-movements can actually become harder to detect, as the device struggles to distinguish between very small changes in position.
“It feels intuitive that sensors tightly fixed to the body would always outperform anything mounted on loose fabric, and that’s what we thought at first.
“Yet for simple tasks, like distinguishing between a small set of postures, the loose fabric sensors performed just as well. This result showed that everyday garments contain far more motion information than we previously thought,” said Dr Howard.
The researchers found that loose fabric effectively acts as a “mechanical amplifier,” making movement easier to detect. When a person moves their arm, for example, a loose sleeve folds, billows, and shifts in complex ways, reacting more sensitively to motion than a tightly fitting sensor. This effect could help bring smart clothing closer to everyday use, with sensors discreetly embedded into items such as shirt buttons as an alternative to bulky devices.
Designing for real-world clothing
Material choice and garment fit were central considerations for the team. Different fabrics influence motion in different ways, but the system is designed to learn patterns across multiple conditions rather than relying on a single type of cloth.
“Different fabrics and fits influence motion, but our framework is designed to accommodate this variation by learning patterns across multiple conditions, which fabric and fit fall under. In some cases, looser fits actually improved classification because the fabric amplified motion cues.”
“In practice, reliability is maintained through calibration and training across a diverse dataset that represents real-world variability. The potential list of clothes, fits, and materials people might wear are endless, so our training data needs to be somewhat representative of this world.”
Beyond healthcare, the researchers believe the findings could influence robotics research and automated technologies that rely on gesture-based control, such as systems used to operate lighting or taps. Sensors were tested on a wide range of fabrics, with both human and robot subjects performing different movements. In every comparison, the fabric-based approach detected motion more quickly and accurately than standard sensors attached to straps or tight clothing, while requiring less data. Looser fabric was also better at distinguishing between very subtle movements.
Co-author Dr Irene Di Giulio, Senior Lecturer in Anatomy and Biomechanics at King’s College London, said: “Sometimes, a patient’s movements are too small for a tight wristband to catch, and therefore we can’t always get the most accurate data on how conditions like Parkinson’s are affecting people’s everyday lives.
“This breakthrough means we could track people in the comfort of their own homes or a care home, in their everyday clothing. It could become easier for doctors to monitor their patients, as well as medical researchers to gather vital data needed to inform our understanding of these conditions, and develop new therapies, including wearable technologies that cater for these kinds of disabilities.”
Implications for robotics
For Dr Howard, who is an expert in robotics research, the work also opens up new possibilities for understanding human movement at scale. Much of robotics research depends on learning from human behaviour, yet collecting large volumes of realistic movement data remains difficult.
“A lot of robotics research is about learning from human behaviour for robots to mimic, but to do this you need huge amounts of data collected from everyday human movements, and not many people are willing to strap up in a Lycra suit and go about their daily business,” he said.
“This research offers the possibility of attaching discreet sensors to everyday clothing, so we can start to collect the Internet-scale of human behaviour data, needed to revolutionise the field of robotics.”
Long-term monitoring
Dr Howard noted that people are far more likely to wear something that feels like normal clothing than a device that feels medical. This makes long-term data collection in real home environments more feasible, particularly for rehabilitation and early detection of mobility decline.
“This makes it feasible to collect consistent, long-term data in real home environments. This is especially valuable for rehabilitation and the early detection of mobility decline, especially in older patients who may be living alone and find it challenging to come into monitored environments like hospitals.
“This technology creates a stepping stone to the continuous monitoring vital for rehabilitation or chronic condition tracking without needing users to change their daily routines.”
While comfort and familiarity are clear advantages, Dr Howard also emphasised the importance of data security and responsible use, particularly when dealing with biometric information.
“[T]he sensors track movement rather than images, the privacy risks are lower than with camera-based systems. But, clear guidelines and transparency will still be essential for everyday adoption.”
What’s next?
The next step for the team is to validate the system at a larger scale, involving a broader range of users and clothing types. Work is also underway to integrate the sensing elements into durable, washable textile platforms that can be used as normal clothing, a process the researchers anticipate could take up to five years.
“The greatest impact we’re likely to see will be in areas where comfort and long-term wearability will be vital, so at-home rehabilitation, care for the elderly in residential settings and preventive health monitoring in the community are primary use cases.
“Manufacturing robust textile-integrated electronics capable of withstanding the pressures of everyday use remains a challenge, as does deepening our understanding of how cloth encodes human motion.
“But when we overcome this, clothing will become an unobtrusive health monitoring platform – to the benefit of the healthcare system.”
