Wearables in Research: How Consumer Gadgets Are Changing ...

Quick Take

2,930 studies with wearables have been registered in ClinicalTrials.gov since 2012

Our FlexTail sensor was validated in a TU Braunschweig study and showed 60% fewer uncomfortable spine positions in digital construction

Consumer wearables like the Apple Watch show 15.79% deviation in clinical parameters

Remote Clinical Trials and continuous monitoring are becoming the new standard

The boundary between lifestyle gadgets and research instruments is increasingly blurring

Imagine: You enter a construction site. Instead of heavy buckets and endless bending work, a robotic arm controls the concrete supply. You monitor the process – and a thin, flexible sensor on your back continuously measures how much your spine is being loaded. In real time.

This is not a vision of the future. This is a study from TU Braunschweig from January 2026.

The sensor is called FlexTail. We developed it. It comes from Germany. And it represents a fundamental shift: Wearables have made the leap from fitness tracker to validated research instrument.

The numbers are impressive. In 2021, over 533 million wearables were sold worldwide. But the real change is happening where no one would have expected it first: In scientific research.

The question is no longer WHETHER we use this technology in research. The question is: How do we use it correctly? And what pitfalls do we need to avoid?

The Science Speaks: What the Studies Show

The Big Trend: 2,930 Studies Can't Be Wrong

Ito and Miyakoshi (2024) systematically scoured ClinicalTrials.gov. Their result is clear: 2,930 clinical studies use wearables – with particularly explosive growth in the last three years before 2022.

The researchers specifically looked for studies using ActiGraph, Apple Watch, Empatica, Fitbit, or Garmin. They initially found 3,214 entries. After removing duplicates, 2,930 relevant studies remained.

The distribution across disease domains is revealing. Cancer research and cardiovascular diseases dominate the field with approximately 20 percent each of all wearable studies. The remaining 60 percent is distributed across various other areas. Previously, ActiGraph and Fitbit dominated the field. Today, Apple Watch, Empatica, and Garmin are catching up massively. Particularly interesting is that 114 studies use multiple devices simultaneously – a clear sign that researchers want to combine the different strengths of various systems.

Ito and Miyakoshi note in the Clinical Trials Journal: "Observational studies outnumbered intervention studies." Translated, this means: We observe more than we intervene. Pure data collection currently takes precedence over therapeutic intervention.

This is strategically important. Because first we must understand what happens in reality before we try to influence it. And this is exactly where wearables unfold their greatest strength: They show everyday life, not the artificial laboratory state.

Our FlexTail in Action: Human-Robot Collaboration on the Construction Site

The TU Braunschweig study by Sawicki et al. (2026), published in the Springer journal Construction Robotics, is a prime example of research according to Industry 5.0 principles. Two teams of two construction workers each produced identical concrete components – once using traditional casting methods, once with digital support (Shotcrete 3D Printing, or SC3DP).

We developed our FlexTail sensor specifically for this type of application. The system continuously measured spine position and load across all 18 sensor pairs, combined with heart rate and blood lactate concentration, plus subjective assessment via Borg Scale and NASA-Task Load Index.

The researchers wanted to know: How does digital support change psychophysiological strain? And how does this strain relate to productivity?

The results are concrete and measurable. In the digital process with SC3DP, there was a 60 percent reduction in uncomfortable spine positions. Carried weight decreased by 44 percent, distance covered by 37 percent. Perceived exertion decreased by 63 percent, while perceived performance increased by 21 percent. Productivity tripled – a factor of 3 in direct comparison to traditional casting methods.

Interestingly, heart rate and blood lactate remained the same between both methods. The classic laboratory parameters did not capture what our FlexTail made visible: The continuous spinal load throughout the entire workday.

This is the crucial point. Wearables capture nuances and progressions that traditional spot measurements in the laboratory miss. An EKG at the end of the day doesn't show how often the back was in critical positions. Our sensor technology captures this continuous load – an essential advantage over sporadic measurements.

Validation: FlexTail in Direct Comparison with Video Systems

Walkling et al. (2025) pitted our FlexTail against high-resolution video-based pose estimation. The question: Can wearables achieve the same accuracy as camera systems, which are considered the gold standard?

The researchers' answer is clear: Yes. And our FlexTail has additional advantages.

The experimental setup included 10 healthy subjects (8 male, 2 female) performing 11 different activities of daily living (ADLs). Simultaneous recording with FlexTail and OpenPose video analysis was performed, with time windows of only 1 second for near real-time classification.

The technical specifications of our FlexTail: 18 sensor pairs on a flexible strip enable a resolution of 1° per segment. The sampling rate is adjustable from 1-100 Hz. Data transmission runs via Bluetooth Low Energy. Particularly important: The wearable is worn completely under clothing – a decisive advantage for acceptance among subjects.

The results in direct comparison show that our FlexTail with F1-scores of 0.88-0.91 was practically on par with the video method (0.90-0.93). Our wearable has decisive practical advantages: It works completely under clothing without visibility needed. It measures 24/7 without elaborate camera setup. It delivers data from real life, not the laboratory. And it causes no privacy issues from cameras.

"The FlexTail recognizes everyday activities just as well as video measurements," write the authors in Sensors. "And is much more suitable for everyday use because it can be worn under clothing."

18 sensor pairs on a flexible strip. A resolution of 1° per segment. This is no longer prototype technology. This is validated, published research equipment – developed by us at MinkTec.

New Study Designs Through Wearables

Continuous Data Collection Instead of Quarterly Medicine

Previously: You go to the doctor, measure blood pressure for 30 seconds, go home. One data point per quarter. Maybe two, if you have a concerned doctor.

Today: A wearable captures 10,000 data points per day. Heart rhythm, activity, sleep quality, movement patterns – around the clock, for weeks and months.

This enables Real-World Evidence on a completely new level. We no longer see how patients behave in artificial laboratory situations. We see how they actually live. We recognize patterns that remain invisible in individual measurements.

An example from research: In gout patients, studies showed that the average number of steps per day decreases by 841 steps during a disease flare. Such connections can only be discovered through continuous monitoring.

Remote Clinical Trials

Decentralized studies are increasingly becoming the standard. Subjects no longer need to travel to the institute weekly to fill out questionnaires or have blood drawn. The data comes automatically, continuously, from real life.

The advantages are measurable. Higher participant numbers become possible because geographic barriers fall. The diversity of participants improves because not only academics from big cities participate. Drop-out rates decrease because the effort for subjects is lower. Study costs are significantly reduced because less staff and fewer rooms are needed.

According to IQVIA (2024), the market for Virtual Clinical Trials is growing from USD 9.27 billion (2024) to USD 15.27 billion (2033). That's growth of almost 65% in nine years.

Biopharma companies are increasingly using wearables in drug studies to better understand the effects of their products. The FDA and EMA accept Real-World Data from wearables in approval procedures – if the data quality is right.

Behavioral Interventions in Real Time

The combination of wearables and apps enables a new quality of intervention: Reaction in real time to behavior.

Possible scenarios are diverse. If heart rate variability drops below a threshold, the app suggests a 2-minute breathing exercise. If activity level remains too low for 3 hours, a notification reminds you to take a walk. If sleep quality is poor for three nights in a row, behavioral tips for bedtime are activated. If an irregular heart rhythm is detected, an automatic warning goes to the treating physician.

This is called Just-in-Time-Adaptive Intervention. The right impulse at the right moment – not generic for everyone, but personalized based on individual data. Researchers also call these "digital biomarkers" – digital biomarkers that can detect changes earlier than traditional methods.

Opportunities and Challenges Considered Honestly

The Opportunities

The benefits of wearables in research emerge from multiple dimensions. Cost efficiency comes from fewer clinical examination visits and reduced staff on site. Data quality improves because thousands of data points replace single measurements – statistically more meaningful, with patterns that remain invisible in spot measurements. Real-time feedback enables immediate reaction to abnormalities, without waiting for the next quarterly appointment. Patient-centricity means data collection in the natural living environment, not in artificial laboratory situations that distort behavior. Scalability allows large-scale studies with moderate budgets because geography is no longer a limiting factor.

What to Consider

Data Quality: The Difference Makes the Success

The quality of wearable data varies considerably depending on the device and application. A comprehensive meta-analysis of various studies shows the full spectrum: Caserman et al. (2024) found deviations of up to 21.47% in VO2max for the Apple Watch Series 7 in athletic subjects. At the same time, Mendt et al. (2025) showed in their laboratory study that the Withings Pulse HR achieved very good agreement with research devices at low activity – with a correlation of r ≥ 0.82 and a deviation of only 3.1 beats per minute. Sen-Gupta et al. (2019) validated the BioStamp nPoint System with outstanding results: 0.957 correlation for heart rate and 0.965 for heart rate variability compared to FDA-cleared reference devices.

The decisive factor is the appropriate device selection for the respective application. While complex metabolic parameters such as VO2max place high demands, many modern wearables already deliver impressively accurate data for heart rate, steps, or sleep patterns. The research shows: With the right device selection and validated algorithms, consumer wearables can indeed achieve scientific quality.

Development Areas with Great Potential

The standardization of wearable data is an active research field that is making rapid progress. While individual manufacturers currently still use proprietary algorithms, the community is increasingly developing open standards and interoperable formats. Regulatory frameworks are also becoming clearer – the FDA and EMA have already published guidelines for the use of Real-World Data from wearables in approval procedures.

The data flood of 10,000 data points per day is both an opportunity and a challenge. Modern machine learning algorithms can efficiently filter and analyze these data volumes. Liu et al. (2023) impressively demonstrated how XGBoost models with wearable data achieve a prediction accuracy of 96% (AUROC) for critical health events – an example of the potential of intelligent data analysis.

Long-term studies require well-designed study protocols to maintain adherence over twelve months. Promising approaches are emerging here: Gamification elements, personalized feedback systems, and regular check-ins can significantly improve long-term use. The experience from over 2,930 published studies provides valuable best practices here.

Where Do We Stand in 2025?

Wearables are not a replacement for traditional scientific methods. They are a useful addition – when used correctly.

Success depends on thorough validation of the chosen system, methodological rigor in study design, realistic expectations of data quality, and sufficient resources for data analysis.

The market is maturing. Regulatory frameworks are becoming sharper and clearer. But: Not every portable gadget is automatically research-ready. The difference between consumer-grade and medical-grade remains relevant.

What Does This Mean for Your Research?

Practical Recommendations for Researchers

For device selection, you should critically check whether CE marking is available, whether published validation studies exist for your target parameters, which parameters were validated how exactly (F1-Score, ICC, MAPE), and whether the device fits your target group (age, fitness level, disease).

Study design requires careful planning. Primary and secondary endpoints must be clearly defined. Sample size should be calculated – wearables do not automatically mean a larger sample is needed. A control group should be established (placebo wearable or standard-care). Drop-out rates must be realistically estimated.

Data management should be professionally established with secure, GDPR-compliant data storage, backup strategies for technical failures, quality control of data streams (outliers, mismeasurements), and documentation of data processing for audit security.

Ethics and data protection must not be forgotten. Informed consent for continuous monitoring must be obtained. Data protection aspects should be transparently communicated. Opt-out options must be guaranteed at all times. Data deletion after study completion should be planned.

Budget should be realistically calculated including device costs (with reserve for defects and losses), data evaluation (software licenses, qualified personnel), support for subjects (tech hotline, training), and additional effort for data management and security.

MinkTec as Partner for Wearables Studies

We developed the FlexTail sensor, which was validated in the TU Braunschweig study under real construction site conditions. Not in the lab. Not simulated. Under dust, noise, and real physical strain.

Application areas of our FlexTail include construction site and production (ergonomics, load measurement), healthcare (rehab monitoring, spine analysis), research (movement analysis, activity recognition), and sports science (movement analysis, posture correction).

What we offer at MinkTec: Technical consulting for study design, sensor configuration for your specific use case, support in data analysis and interpretation, and experience from validated research projects.

Are you planning a study with wearables? We accompany you from the first idea through study design to data analysis and publication.

Conclusion and Outlook

The numbers speak a clear language: 2,930 studies with wearables since 2012 in ClinicalTrials.gov. 60% reduction in uncomfortable spine positions (TU Braunschweig study with our FlexTail). F1-Score of 0.90 for FlexTail (validated against professional video systems). 15.79% deviation in consumer wearables (warning against uncritical adoption). USD 35.1 billion market volume for clinical-grade wearables 2024.

The boundary between lifestyle gadget and scientific research instrument is increasingly blurring. But: Just because something is portable doesn't automatically make it scientifically valid.

The crucial questions before your next project: Is my chosen wearable scientifically validated for the planned parameters? Are my research questions better answered through continuous data collection? Have I budgeted for personnel for meaningful data analysis? Do I know the regulatory requirements (GDPR, FDA, ethics committee)?

Those who test now gain valuable experience with Real-World Evidence. Those who wait miss the connection to a standard that is rapidly establishing itself.

German technology like our FlexTail shows: We don't have to import. We can ourselves define what research maturity means in wearable technology – and deliver the scientific evidence for it.

Want to know if our FlexTail fits your research project? Contact us for a non-binding initial consultation. We'll show you concrete application examples from validated studies and help you with study design.

[Schedule Consultation Now]

References

Sawicki, B., Düking, P., Placzek, G., Masur, L., Dörrie, R., Schwerdtner, P., & Kloft, H. (2026). Human–robot collaboration in digital fabrication with concrete: quantifying productivity and psychophysiological strain of human workers. Construction Robotics, 10(4). https://doi.org/10.1007/s41693-025-00173-x

Ito, Y. M., & Miyakoshi, T. (2024). Assessing the current utilization status of wearable devices in clinical research. Clinical Trials, 21(3), 470-482. https://doi.org/10.1177/17407745241230287

Walkling, J., Sander, L., Masch, A., & Deserno, T. M. (2025). Wearable Spine Tracker vs. Video-Based Pose Estimation for Human Activity Recognition. Sensors, 25(12), 3806. https://doi.org/10.3390/s25123806

Caserman, P., Yum, S., Göbel, S., Reif, A., & Matura, S. (2024). Assessing the accuracy of smartwatch-based estimation of maximum oxygen uptake using the Apple Watch Series 7: Validation study. JMIR Biomedical Engineering, 9, e59459. https://doi.org/10.2196/59459

IQVIA Institute for Human Data Science. (2024). Digital Health Trends 2024. https://www.iqvia.com/insights/the-iqvia-institute/reports-and-publications/reports/digital-health-trends-2024

Mendt, S., Zout, G., Rabuffetti, M., Gunga, H., Bunker, A., Barteit, S., & Maggioni, M. (2025). Laboratory comparison of consumer-grade and research-established wearables for monitoring heart rate, body temperature, and physical activity in sub-Saharan Africa. Frontiers in Physiology, 16, 1491401. https://doi.org/10.3389/fphys.2025.1491401

Sen-Gupta, E., Wright, D. E., Caccese, J. W., Wright, J. A., Jortberg, E., Bhatkar, V., Ceruolo, M., Ghaffari, R., Clason, D., Maynard, J. P., & Combs, A. (2019). A pivotal study to validate the performance of a novel wearable sensor and system for biometric monitoring in clinical and remote environments. Annals of Biomedical Engineering, 47(9), 1934-1946. https://doi.org/10.1159/000493642

Liu, J. H., Shih, C. Y., Huang, H. L., Peng, J. K., Cheng, S. Y., Tsai, J. S., & Lai, F. (2023). Evaluating the potential of machine learning and wearable devices in end-of-life care in predicting 7-day death events among patients with terminal cancer: Cohort study. Journal of Medical Internet Research, 25, e47366. https://doi.org/10.2196/47366

Article written on February 2, 2026. All cited studies were fully read before publication and checked for their relevance to the topic.

Wearables in Research: How Consumer Gadgets Are Changing Science