Ambient Intelligence · Methodology · Naming Taxonomy
Naming Taxonomy
Ambient Intelligence
Passive health monitoring, derived from the environment.
Ambient Intelligence is a passive, sensor-based health monitoring platform that derives clinically meaningful indices from ambient lidar and radar signals — no wearables required. Where traditional digital health relies on what a person chooses to wear, Ambient Intelligence measures what a space observes: movement, gait, sleep, sedentary behavior, circadian rhythm, and fall risk, all inferred from point-cloud and height-signal data captured continuously in the background of daily life.
Naming Taxonomy
Tier 1 · Platform
Ambient Intelligence
One global namespace.
All composite indices derive from this root.
All composite indices derive from this root.
Tier 2 · Domain Indices
→ algorithmlab for interactive explorationAmbientActivityIndex
Aggregate movement score derived from per-frame point-cloud activity; the ambient parallel to accelerometer-based activity counts.
→ algorithmlab
AmbientSleepIndex
Composite sleep quality score combining architecture, fragmentation, and HRV-derived recovery.
→ algorithmlab
AmbientGaitIndex
Gait quality composite from step detection, cadence, gait speed, and step-time variability.
→ algorithmlab
AmbientSedentaryIndex
Sedentary burden measure combining bout analysis, break frequency, and intensity profile.
→ algorithmlab
AmbientRecoveryIndex
Recovery composite from HRV, sleep, and circadian inputs.
→ algorithmlab
AmbientCircadianIndex
Rhythm regularity score rolling up IS, IV, RA, and M10/L5.
→ algorithmlab
AmbientMetabolicIndex
Metabolic composite derived from CGM trace, time-in-range, and variability.
→ algorithmlab
AmbientRiskIndex
Fall risk composite driven by SHAP-attributed features across gait, activity, and circadian domains.
→ algorithmlab
Tier 3 · Component Metrics
— ASTP · SATP · IS · IV · RA · M10/L5 · AGRU · SHAP · cadence · gait speed · step detection · step-time variability · TUG · METs · breaks/hour · LF/HF · Poincaré · stress index · TIR · AGP · CV% · MAGE · sleep architecture · SII · HRV · recovery · chronotype · fragmentation —
Tier 4 · Raw Signal Fields
Algorithm Mathematics
Foundation Layer
Signal Processing
Clinical
significance
significance
Smoothing, normalization, and frequency-domain decomposition are the prerequisite layer for every downstream index. Without them, motion artifact and sensor noise contaminate every higher metric — DFA, SampEn, gait speed, and fall thresholds all assume a denoised input. This is the foundation that makes the rest of the system clinically usable.
Simple Moving Average (SMA)
definition
SMA(t, w) = 1w · Σi=t−w+1t xi
tcurrent time index
wwindow size (epochs)
xᵢsignal value at index i
Exponential Smoothing
Single exponential (Holt 1957). α controls the memory decay rate — higher α tracks signal more closely, lower α produces stronger smoothing.
recursive update
St = α · xt + (1 − α) · St−1, S0 = x0, α ∈ (0, 1]
αsmoothing factor
Sₜsmoothed estimate at t
xₜobserved signal at t
Rolling Z-Score
Detects anomalies by measuring how many standard deviations a reading falls from its local rolling mean. |z| > threshold flags an anomaly.
standardisation
zt = xt − μ̂tσ̂t
μ̂t = 1w · Σi=t−w+1t xi, σ̂t = √1w · Σi=t−w+1t (xi − μ̂t)2
Autocorrelation Function (ACF)
Measures the self-similarity of a signal at lag k. Periodic signals produce peaks at their fundamental frequency and harmonics.
normalised lag-k ACF
rk = Σt=1n−k (xt − x̄)(xt+k − x̄)Σt=1n (xt − x̄)2, k = 0, 1, 2, …
AmbientActivityIndex · AmbientRiskIndex
Nonlinear Complexity Theory
AmbientCircadianIndex · AmbientActivityIndex
Circadian Rhythm · Fragmentation
AmbientGaitIndex
Gait Analysis
AmbientActivityIndex
Metabolic Equivalents · Energy Expenditure
AmbientSleepIndex
Sleep Architecture · HRV
AmbientRecoveryIndex · AmbientCircadianIndex
Recovery · Chronotype
AmbientSedentaryIndex
Sedentary Behavior
AmbientFallDetection
Fall Detection
AmbientMetabolicIndex
Glycemic Analysis · CGM
AmbientRecoveryIndex
Autonomic Nervous System · HRV Frequency Domain
AmbientRiskIndex
Predictive Risk · Anomaly Detection
AmbientActivityCounts · AmbientActivityIndex
Radar · mmWave Point Cloud
Academic References
Signal Processing
Holt (1957) · Forecasting trends and seasonals by exponentially weighted moving averages — ONR Memorandum 52
Box & Jenkins (1970) · Time Series Analysis: Forecasting and Control — autocorrelation foundations
Nonlinear Complexity
Peng et al. (1994) · Mosaic organization of DNA nucleotides — Phys. Rev. E 49(2): DFA introduction
Richman & Moorman (2000) · Physiological time-series analysis using approximate and sample entropy — Am. J. Physiol. 278: H2039–H2049
Bandt & Pompe (2002) · Permutation entropy: a natural complexity measure for time series — Phys. Rev. Lett. 88(17)
Hurst (1951) · Long-term storage capacity of reservoirs — Trans. Am. Soc. Civ. Eng. 116: 770–808
Costa, Goldberger, Peng (2002) · Multiscale entropy analysis of complex physiologic time series — Phys. Rev. Lett. 89(6)
Circadian / Activity
Witting et al. (1990) · Alterations in the circadian rest-activity rhythm in aging and Alzheimer's disease — Biol. Psychiatry 27(6)
Roenneberg et al. (2003) · Life between clocks: daily temporal patterns of human chronotypes — J. Biol. Rhythms 18(1)
Gait
Studenski et al. (2011) · Gait speed and survival in older adults — JAMA 305(1): 50–58
Hausdorff et al. (2001) · Gait variability and basal ganglia disorders — Mov. Disord. 13(3)
Podsiadlo & Richardson (1991) · The Timed Up & Go: a test of basic functional mobility — JAGS 39(2)
METs
Ainsworth et al. (2011) · Compendium of Physical Activities: 2011 update — Med. Sci. Sports Exerc. 43(8)
WHO (2020) · Guidelines on physical activity and sedentary behaviour — Geneva, World Health Organization
Sleep / HRV
Berry et al. (2012) · AASM Manual for the Scoring of Sleep and Associated Events — American Academy of Sleep Medicine
Task Force ESC/NASPE (1996) · Heart rate variability: standards of measurement — Circulation 93(5)
Lichstein et al. (2006) · Quantitative criteria for insomnia — Behav. Res. Ther. 41(4)
Sedentary Behavior
Matthews et al. (2012) · Amount of time spent in sedentary behaviors in the US, 2003–2004 — Am. J. Epidemiol. 167(7)
Healy et al. (2008) · Breaks in sedentary time: beneficial associations with metabolic risk — Diabetes Care 31(4)
Fall Detection
Tinetti et al. (1988) · Risk factors for falls among elderly persons living in the community — NEJM 319(26)
Lin et al. (2026) · AGRU: attention-gated recurrent unit for multi-modal fall risk classification
CGM / Metabolic
Bergenstal et al. (2019) · Time-in-range as a glycemic outcome metric — Diabetes Care 42(10)
Service et al. (1970) · Mean amplitude of glycemic excursions, a measure of diabetic instability — Diabetes 19(9)
Autonomic HRV
Akselrod et al. (1981) · Power spectrum analysis of heart rate fluctuation — Science 213(4504): 220–222
Brennan, Palaniswami, Kamen (2001) · Do existing measures of Poincaré plot geometry reflect nonlinear features — IEEE Trans. Biomed. Eng. 48(11)
Predictive Risk
Page (1954) · Continuous inspection schemes — Biometrika 41(1/2): 100–115 (CUSUM)
Takens (1981) · Detecting strange attractors in turbulence — Lecture Notes in Mathematics 898
Lundberg & Lee (2017) · A unified approach to interpreting model predictions — NeurIPS 2017 (SHAP)
Radar / mmWave
Texas Instruments (2024) · IWR6843AOP single-chip 60-GHz FMCW radar sensor — Technical Reference Manual
Adib et al. (2015) · Smart homes that monitor breathing and heart rate — MIT CSAIL (RF-Pose foundations)
Zhao et al. (2020) · Through-wall human pose estimation using radio signals — MIT CSAIL
Selected canonical citations · full bibliography available on request
Market Positioning
Apple Health has Cardio Fitness. Oura has Readiness. Whoop has Strain. Ambient Intelligence has a family of Ambient Indices — the same clinical-grade composites, derived passively from the environment instead of the wrist.
Apple Health
Cardio Fitness
Oura
Readiness
Whoop
Strain
Ambient
Ambient Indices
— Ambient Intelligence · Methodology · AmbientActivityIndex · AmbientSleepIndex · AmbientGaitIndex · AmbientSedentaryIndex · AmbientRecoveryIndex · AmbientCircadianIndex · AmbientMetabolicIndex · AmbientRiskIndex —