Model

UMP

Unified Model of Performance (WRAIR/BHSAI, Reifman group)

Cite
UnconfirmedCatalog paper date: 2016Added to atlas: Nov 9, 2026

Activity counts measure attention, not model quality. Missing counts are unknown. Publication status follows the linked source; a listed venue alone does not confirm peer review.

The Unified Model of Performance (WRAIR/BHSAI, Reifman group) predicts group-average neurobehavioral performance from sleep loss and caffeine by multiplying a dose-dependent caffeine factor onto caffeine-free performance. It was assessed in 14 laboratory and field conditions, covering seven sleep-loss schedules (from 5 h sleep each night to 85 h continuous wake) and six caffeine doses (placebo to repeated 200 mg or a single 600 mg dose), and predicted group-average PVT performance across those conditions (10.5665/sleep.6164).

Signals
caffeine dose/time, sleep/wake timing
Form factor
software (sleep/schedule input)
License
Published model (US Government work)
Year
2016
Architecture

Two-process-derived model extended to chronic sleep restriction, with a multiplicative dose-dependent caffeine factor (2016); Bayesian individualisation in 2B-Alert

At a glance
Type
Task model
Form factor
Software (sleep/schedule input)
Parameters
Not reported
Architecture
Two-process-derived model extended to chronic sleep restriction, with a multiplicative dose-dependent caffeine factor (2016); Bayesian individualisation in 2B-Alert
Venue
Not reported
Year
2016
arXiv
Not reported
DOI
Not reported
Code
Open
Weights
Not released
Licence
Published model (US Government work)
Caveats
Not reported
Datasets
Not reported
Added
Nov 9, 2026
Updated
Nov 9, 2026
UMP: ChatPPG research summary card

How it works

How UMP works: Input signals → Model → Downstream tasksInput signals: caffeine dose/time; sleep/wake timing. Model: UMP. Downstream tasks: Stress.Input signalscaffeine dose/timecaffeine dose/timesleep/wake timingsleep/wake timingModelUMPDownstream tasksStressStress
Architecture
Two-process-derived model extended to chronic sleep restriction, with a multiplicative dose-dependent caffeine factor (2016); Bayesian individualisation in 2B-Alert
Inputs
caffeine dose/time, sleep/wake timing
Approach
fatigue · biomathematical · classical

Datasets used

No public datasets linked yet.

Papers· 4

Cite this model

Cite the publication recorded for this model. Check it against the publisher before submitting.

UMP. (2016).
@misc{unifiedmodelofpe2016,
  title         = {{UMP}},
  year          = {2016}
}

Source metadata: 2016

Record updated Nov 9, 2026

History

  1. Added to the ChatPPG research hub