Duration
Beyond the laboratory window
The 80 hours span three full nights. From hour 48 onwards, the record enters sparsely described territory: most of the total sleep-deprivation literature stops before that point.
An educational environment running for 80 continuous hours, fully monitored, with a subject undergoing prolonged sleep deprivation under sustained cognitive, vocal and postural effort.
This page describes what will be measured during the world record attempt of The Longest Lesson in History, with which instruments, at what frequency, and which correlations the research intends to investigate afterwards. The purpose of the collection is not operational: it is to produce a dense, consented and auditable time series for later scientific study.
Laboratory sleep-deprivation studies rarely exceed 36 to 48 hours, and almost never involve the subject performing a complex, public, communicatively demanding cognitive task. Here, the same individual spends 80 hours in continuous teaching activity, speaking almost the entire time, seated and standing, on camera, with a room full of students depending on his performance.
The monitoring exists first and foremost for safety. But the same instrumentation that protects also produces, as a by-product, a dataset that cannot be assembled under normal conditions: high temporal density, multiple domains measured in parallel and a baseline captured months in advance.
The 80 hours span three full nights. From hour 48 onwards, the record enters sparsely described territory: most of the total sleep-deprivation literature stops before that point.
The subject is not merely awake: he is teaching. Every hour has a learning objective, a practical activity and a measurement of class performance, which gives an external, objective measure of fatigue's effect on the task.
The instructor has type 2 diabetes mellitus managed with medication. The cognitive degradation curve and glycaemic response of a subject in this condition across 80 hours of wakefulness is essentially undescribed in the literature.
Baseline indicators measured months in advance, plus progressive 24- and 36-hour wakefulness simulations, allow every point in the series to be compared with the subject himself rather than with a population average.
Besides the instructor there is an in-person cohort at individual workstations, with named entry and exit control and performance measured hour by hour, allowing the monitored subject to be compared with people going through the same night.
No line of enquiry may alter the pedagogical design, compromise safety or interfere with the compliance of the attempt. Questions that would require experimental manipulation are recorded for future studies.
The table consolidates the indicators set out in the health protocol. The highlighted rows are the continuously captured ones: those that produce a dense time series rather than isolated points, and on which almost all of the correlation analysis described below depends.
| Indicator | Category | Frequency | Method |
|---|---|---|---|
| Heart rate (HR) | Cardiovascular | Continuous | Wrist wearable + chest strapcontinuous capture |
| Heart rate variability (HRV) | Cardiovascular | Continuous, with structured measurement every 6 h | Wrist and finger wearablescontinuous capture |
| Oxygen saturation (SpO₂) | Cardiovascular | Every 6 h plus continuous sampling | Wearable + finger oximetercontinuous capture |
| Peripheral temperature | Clinical | Continuous · logged every 12 h | Finger wearable + digital thermometercontinuous capture |
| Actigraphy and movement | Motor | Continuous | Wearable accelerometrycontinuous capture |
| Sleep and microsleep detection | Neurological | Continuous | Wearable + manual event markingcontinuous capture |
| Autonomic stress index | Cardiovascular | Continuous | Derived from HRV by the wearablecontinuous capture |
| Blood pressure | Cardiovascular | Every 6 h up to 48 h; every 3 h thereafter | Digital sphygmomanometer |
| Blood glucose | Metabolic | Continuous | Continuous glucose monitor (CGM)continuous capture |
| Body weight | Metabolic | Every 24 h | Bioimpedance scale |
| Lower-limb oedema | Cardiovascular | Every 12 h | Structured visual assessment |
| KSS · Karolinska Sleepiness Scale | Neurological | Every 3 h | Questionnaire |
| PVT · Psychomotor Vigilance Test | Neurological | Every 3 h | App on a dedicated tablet |
| Stroop Test | Neurological | Every 6 h up to 48 h | Cognitive test |
| Trail Making Test A/B | Neurological | Every 6 h up to 48 h | Cognitive test |
| Fatigue scale (0–10) | General | Every 3 h | Self-assessment |
| Mood, motivation and stress (0–10) | Psychological | Every 6 h | Cognitive-behavioural check-in |
| Neck, lower-back and lower-limb pain (0–10) | Musculoskeletal | Every 6 h | Self-assessment |
| Hoarseness and vocal effort | Vocal | Every 6 h | Perceptual assessment + acoustic measures |
| Eye dryness and fatigue | Visual | Every 6 h | Subjective assessment + log of 20-20-20 breaks |
Seven of the twenty indicators depend on continuous capture by a wearable sensor. The rest are point measurements taken in the 3-, 6-, 12- or 24-hour windows.
The difference between a clinical safety record and a research dataset lies in sampling density. Measurements every six hours describe a trend; wearable sensors describe a trajectory, and it is the trajectory that makes it possible to answer whether one marker precedes another.
An uninterrupted series of HR and HRV across the 80 hours, with enough resolution to observe the response to caffeine micro-doses and to the more complex content blocks. HRV is the study's leading candidate for an early marker of fatigue.
The subject cannot stop to be measured: the lesson is continuous and interrupting it would end the attempt. Any measurement requiring active attention competes with the task; passive capture is the only kind that runs across the full 80 hours.
SpO₂ and peripheral temperature across prolonged wakefulness and sustained seated posture; accelerometry to describe the pattern of spontaneous movement and adherence to the postural rotation protocol.
The three scheduled 60-minute breaks are the only rest periods in the attempt. Knowing what actually happened in each one, and whether effective sleep occurred, is a precondition for measuring the magnitude of the cognitive recovery that follows.
Recording the same signal at two anatomical points creates a second layer of analysis: agreement between devices across 80 hours of uninterrupted wear, with sweat, constant typing and ambient temperature variation. That is a research question in its own right, and it has not been described at this duration.
To be cross-referenced with the lesson log, the questionnaires and the cognitive tests, the series has to leave the device with its granularity and timestamps intact. Without an export aligned to the attempt's official clock, the data does not enter the analysis.
A gap in the time series is not an operational detail: it is the loss of exactly the stretch you set out to study. The critical window — hours 48 to 80, when cognitive degradation sharpens — is also the furthest from any recharge. The same applies to comfort: any device that has to be removed because of discomfort during the task stops recording at precisely the point of greatest interest.
The panels below are the recording structure already defined in the protocol. The values shown are simulated: reference projections derived from the sleep-deprivation literature, used to validate the collection format. They will be replaced by the real values measured during the challenge.
Cardiovascular, thermal and weight parameters across the 80 hours.
| Hour | HR bpm | HRV ms | BP mmHg | SpO₂ % | Temp. °C | Weight kg | Note |
|---|---|---|---|---|---|---|---|
| 0 | 62 | 48 | 118/78 | 98 | 36,4 | 80,0 | Initial baseline |
| 6 | 65 | 46 | 120/80 | 98 | 36,5 | — | No relevant changes |
| 12 | 68 | 44 | 122/80 | 98 | 36,6 | — | Expected physiological response |
| 24 | 72 | 40 | 124/82 | 97 | 36,7 | 79,8 | Early sleep deprivation |
| 36 | 78 | 35 | 128/84 | 97 | 36,8 | — | Slight increase in physiological stress |
| 48 | 84 | 30 | 132/86 | 97 | 37,0 | 79,5 | Moderate fatigue |
| 60 | 92 | 24 | 136/88 | 96 | 37,1 | — | Monitoring intensified |
| 72 | 98 | 20 | 140/90 | 96 | 37,2 | 79,2 | High physiological load |
| 80 | 88 | 28 | 132/84 | 97 | 36,9 | 79,0 | Partial recovery after the final rest period |
Subjective sleepiness, psychomotor vigilance, inhibitory control and cognitive flexibility.
| Hour | KSS 1–9 | PVT time ms | PVT lapses n | Stroop s | Stroop errors n | Trail A s | Trail B s | Note |
|---|---|---|---|---|---|---|---|---|
| 0 | 2 | 240 | 0 | 35 | 0 | 28 | 62 | Initial baseline |
| 6 | 3 | 255 | 0 | 37 | 0 | 30 | 66 | No relevant change |
| 12 | 4 | 285 | 1 | 41 | 1 | 34 | 74 | Mild fatigue |
| 24 | 5 | 340 | 2 | 48 | 2 | 42 | 92 | Slight slowing |
| 36 | 6 | 420 | 4 | 58 | 4 | 55 | 125 | Moderate fatigue |
| 48 | 7 | 520 | 7 | 72 | 6 | 68 | 160 | High cognitive risk |
| 60 | 8 | 680 | 12 | 95 | 10 | 92 | 220 | Neurological assessment required |
| 72 | 8 | 740 | 15 | 110 | 13 | 105 | 260 | Keep under observation |
| 80 | 7 | 610 | 9 | 88 | 8 | 80 | 190 | Close to the end |
0-to-10 scales from the cognitive-behavioural check-in and acoustic voice measures (dB SPL, jitter and shimmer).
| Hour | Stress | Mood | Focus | Irritab. | Self-efficacy | Vocal effort | Vocal fatigue | Throat | Voice qual. 0–4 | dB | jitter | shimmer |
|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 2 | 8 | 8 | 1 | 8 | 2 | 1 | 1 | 4 | 68 | 0,3 | 1,8 |
| 6 | 3 | 7 | 7 | 2 | 7 | 2 | 2 | 2 | 4 | 69 | 0,3 | 1,9 |
| 12 | 4 | 6 | 6 | 3 | 6 | 3 | 3 | 3 | 3 | 70 | 0,4 | 2,1 |
| 24 | 5 | 5 | 5 | 4 | 5 | 4 | 4 | 4 | 3 | 71 | 0,4 | 2,2 |
| 36 | 6 | 5 | 4 | 5 | 4 | 5 | 5 | 5 | 2 | 72 | 0,5 | 2,4 |
| 48 | 7 | 4 | 4 | 6 | 4 | 6 | 6 | 6 | 2 | 73 | 0,6 | 2,6 |
| 60 | 7 | 4 | 3 | 6 | 3 | 6 | 6 | 6 | 2 | 74 | 0,6 | 2,7 |
| 72 | 6 | 5 | 4 | 5 | 4 | 5 | 5 | 5 | 2 | 73 | 0,5 | 2,5 |
| 80 | 4 | 6 | 5 | 4 | 6 | 4 | 4 | 4 | 3 | 71 | 0,4 | 2,2 |
Visual fatigue, adherence to the 20-20-20 rule and musculoskeletal load during prolonged sitting and standing.
| Hour | Eye fatigue | Dryness | Burning | Blurred vision | 20-20-20 breaks /h | Neck pain | Lower-back pain | Lower-limb pain | Stiffness | Mobility | Note |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 1 | 1 | 1 | 0 | 6 | 1 | 1 | 0 | 1 | 9 | — |
| 6 | 2 | 2 | 1 | 1 | 5 | 2 | 1 | 1 | 2 | 9 | — |
| 12 | 3 | 3 | 2 | 1 | 4 | 3 | 2 | 2 | 3 | 8 | Slight tension |
| 24 | 4 | 4 | 3 | 2 | 4 | 4 | 3 | 3 | 4 | 7 | Muscle tension |
| 36 | 5 | 5 | 4 | 2 | 3 | 5 | 4 | 4 | 5 | 6 | Mild discomfort |
| 48 | 6 | 6 | 5 | 3 | 3 | 6 | 5 | 5 | 6 | 5 | Moderate stiffness |
| 60 | 7 | 6 | 5 | 4 | 2 | 7 | 6 | 6 | 6 | 4 | Muscle fatigue |
| 72 | 7 | 7 | 6 | 4 | 2 | 7 | 6 | 6 | 7 | 4 | High discomfort |
| 80 | 6 | 6 | 5 | 3 | 3 | 6 | 5 | 5 | 5 | 5 | Partial recovery |
On the 0-to-10 scales, unless stated otherwise, 0 is the best condition and 10 the worst. Functional mobility is the exception: 10 is the best condition.
Nine indicators from the three domains, on the same 80-hour timeline, with the three scheduled 60-minute breaks marked. This is the format in which the analysis happens: not in an isolated number, but in the shape of the curve and in what it does before and after each break.
Each panel has its own scale. The gold band marks the scheduled 60-minute breaks, at hours 39, 61 and 74.
Simulated values. The full table for each panel is in the previous section.
Measuring each indicator in isolation describes a body growing tired. The scientific value lies in the cross-referencing: which signal moves first, which merely follows, and which stops following. Below are the pairs that structure the research programme.
If heart rate variability falls before the lapses appear, it works as an early indicator of fatigue, and becomes an actionable signal rather than a merely descriptive one.
There is a point at which subjective sleepiness saturates while objective performance continues to degrade. Identifying that dissociation means identifying the moment when self-assessment stops being reliable.
Each point is a measurement window, labelled by lesson hour. The dashed line connects the points in time order. Simulated data.
The central hypothesis of the cardiovascular block: to check whether the fall in heart rate variability precedes — and by how long — the worsening of reaction time and the increase in lapses on the psychomotor vigilance test.
Sleep deprivation degrades precisely the ability to assess oneself. Determining the hour at which the subjective scale stops tracking the objective measure has direct practical consequences for any safety protocol based on self-report.
With continuous glucose monitoring in a subject with type 2 diabetes on a split-meal nutrition protocol, it becomes possible to check whether glycaemic excursions correlate in time with drops in performance, and whether splitting meals really does reduce post-prandial sleepiness.
Five doses of 80 to 100 mg, with times logged, against the cardiovascular response in the two hours after each. The question is whether the effect holds, weakens or reverses as sleep debt accumulates.
Tests administered immediately before and after each of the three long breaks, to measure the magnitude of the recovery and check whether it diminishes with each repetition. Here the wearable's detection of effective sleep is what distinguishes a genuine rest break from a merely administrative one.
The series is anchored to clock time, not only to lesson hour. That makes it possible to separate the effect of accumulated debt from the effect of time of day, and to check whether there is partial recovery during daylight periods despite the accumulation.
Jitter, shimmer and intensity every six hours, plus speech rate, lexical density and pause frequency extracted from the transcript of the full audio. The question is whether the voice betrays fatigue before the subject himself notices it.
Every hour of the lesson ends with five multiple-choice questions answered individually by the class. That is 385 performance measurements anchored to the same timeline: an external measure of the effect of the instructor's fatigue on the task he is performing.
The same signal, recorded simultaneously at the finger and the wrist, for 80 uninterrupted hours, with constant typing, temperature variation and sweating. The convergence and divergence between the two series are themselves a result, and no record of this comparison exists at this duration.
Neck, lower-back and lower-limb pain scores every six hours, cross-referenced with the log of time spent in each position, with the oedema assessment every twelve hours and with actigraphy, to evaluate the effectiveness of the postural rotation protocol.
The collection described here involves health data belonging to an identified individual, image and voice on open broadcast, and individual participant performance. The processing was designed before the collection, not after it.
Participation, image, assessment and research are consented to separately. Accepting one does not imply accepting the others.
If aggregate data will do, no named identification is collected. Where anonymisation is possible, it is the preferred route. Operational databases and research databases remain separate.
Research involving human subjects is treated as an autonomous strand, with a protocol, named leads and ethical review preceding the collection. The event can exist without the research; the research does not emerge informally inside the event.
The corpus brings together audio, video, biometrics and individual performance, with a defined regime for custody, retention and deletion, and a sharing model that allows it to be opened to the scientific community while preserving the rights of the data subjects.
If a measure greatly increases a participant's exposure and brings little real gain in safety or research value, it does not go in. If it improves safety, reduces uncertainty and can be explained transparently, it deserves a place in the protocol.
Authority to stop: the decision to end the attempt for clinical reasons does not rest with the instructor, because sleep deprivation degrades precisely the capacity for self-assessment. A named healthcare professional holds that authority, with thresholds known to the whole team before the start.
At the end of the 80 hours, what remains is not only a record. It is a time-aligned dataset: a continuous series of physiological signals, twenty-seven cognitive and perceptual measurement windows, continuous glucose monitoring, full time-stamped audio and video, the complete lesson log with complexity and method recorded hour by hour, and 385 class performance measurements.
All anchored to the same clock, with a baseline measured beforehand and re-measurement in the 72 hours after the end. It is that alignment, not the number of sensors, that makes the correlations on this page investigable.
The values presented on this page are simulated and serve solely to demonstrate the structure of the collection and the type of analysis intended. After the challenge, the real data will be used in scientific research, with methodological rigour and research ethics.