Andi Assess

AI-generated draft

Independent living assessment report

General Behavioural and Daily-Living Evidence Report

Across 22 behavioural days the individual showed consistent daily in-home activity centred on the kitchen, with an average of 102.5 room episodes and 38.73 room transitions per behavioural day and no behavioural day without a detected kitchen episode.

Assessment at a glance

RAG shows priority for professional review. Confidence shows the strength of the observation.

Green Moderate confidence

Kitchen and Nutrition-Related Activity

Frequent short kitchen presences occur every behavioural day, clustering into mid-morning, midday and early-evening windows.

Amber Moderate confidence

Bathroom Use

Bathroom use consists of many brief daytime episodes with a midday peak and pronounced day-to-day variability in count.

Green Moderate confidence

Movement and Room Use

Movement is concentrated between 09:00 and 19:00 with twin midday and early-evening peaks, and very low episode counts between 03:00 and 08:00.

Amber Low confidence

Overnight Activity and Quiet Periods

Most nights contain frequent detected activity with short longest-quiet periods, alongside few overnight bathroom episodes.

Amber Low confidence

Routine Consistency

The shape of the active day is moderately repeatable hour-by-hour, but the times of the first and last daily activity vary substantially.

Grey Not-reportable confidence

Property Exits and Community Access

No pattern can be described because exit telemetry is not available for this domain.

Red prompt review · Amber discuss or check · Green no notable adverse change identified · Grey not assessable or not reportable

Service-user reference: SU001
Assessment period: Requested 20 April 2026 to 30 May 2026 (Europe/London); telemetry observed only from 30 April 2026 to 21 May 2026, giving 22 behavioural days of 41 calendar days.
Generated: August 03, 2026

Executive summary

To describe detected home telemetry patterns relevant to general wellbeing for service user SU001, supporting professional judgement without diagnosis, risk determination or causal inference.

The kitchen accounted for 52.02% of all room episodes (1,173 of 2,255) with an average of 85.35 detected occupancy minutes per behavioural day.

Kitchen activity clustered into three recurring windows: 10:00-11:00, 11:00-12:00 and 17:00-18:00.

Bathroom episodes averaged 9.91 per behavioural day with an average duration of 43 seconds, a longest episode of 146 seconds and no episode exceeding ten minutes.

Across 20 analysed nights the average longest quiet period was 57 minutes (median 40, shortest 14), and overnight bathroom episodes averaged 0.6 per night with 12 nights having none.

Day-start and day-end timings varied widely: first-activity and last-activity consistency scores were both 0.0 (standard deviations 293.4 and 577.75 minutes), while hour-profile similarity between consecutive days was 0.4921.

Changes over time

Overall activity was categorised as stable, moving from a baseline average of 101.29 episodes per day (30 April-6 May) to 96.86 (15-21 May), a change of -4.37%.
Bathroom episodes were categorised as decreasing, from 10.71 per day at baseline to 9.29 per day in the comparison window, a change of -13.33%.
Weekly average activity per behavioural day rose from 94.75 (week of 27 April) to 119.0 (week of 11 May) before falling to 81.5 in the final part-week of 18 May, which contained only four behavioural days.

Overall interpretation

Telemetry describes a kitchen-centred daily pattern with regular meal-time clustering, short bathroom episodes and highly variable start and end times to the active day, on a partial dataset covering 22 of 41 requested days.

Assessment finding

Kitchen and Nutrition-Related Activity

Green Moderate confidence

What was detected

Kitchen episodes averaged 53.32 per behavioural day, totalling 1,173 episodes with an average duration of 96 seconds and a median of 54 seconds.
No behavioural day was recorded without a kitchen episode, and detected kitchen occupancy averaged 85.35 minutes per behavioural day.
No appliance-category kitchen sessions were derived (0.0 sessions per behavioural day and 0 weekly appliance episodes in every observed week).

Pattern identified

Frequent short kitchen presences occur every behavioural day, clustering into mid-morning, midday and early-evening windows.

Why this may matter

Frequent short kitchen presences may reflect repeated drink or snack preparation, or repeated passage through a kitchen that also serves as a circulation space; telemetry cannot distinguish these.
The absence of derived appliance sessions may reflect an episode-derivation limitation rather than an absence of cooking, since kitchen active-power telemetry was recorded but produced no reportable sessions.
Meal-window clustering could indicate a settled eating rhythm and should be interpreted alongside any observations of dietary intake or support arrangements.

What to explore

Does the kitchen layout mean it is used as a through-route between other rooms?
What does the individual or their support network describe about meal preparation and frequency of hot drinks?
Is cooking appliance use expected to be captured by the installed power sensors?

What the data cannot tell us Detected presence does not establish that food or drink was prepared or consumed. and Appliance-session metrics returned zero and cannot be used to infer that appliances were unused.

Assessment finding

Bathroom Use

Amber Moderate confidence

What was detected

218 bathroom episodes were recorded, averaging 9.91 per behavioural day with peak episode starts in the 12:00 hour.
Episodes were short, averaging 43 seconds with a median of 37 seconds and a maximum of 146 seconds; none exceeded ten minutes.
Daily counts ranged from 1 on 20 May to 34 on 1 May, with most days between 5 and 14.

Pattern identified

Bathroom use consists of many brief daytime episodes with a midday peak and pronounced day-to-day variability in count.

Why this may matter

Consistently brief episodes may be consistent with quick functional use, or may reflect sensor episode-splitting; telemetry alone cannot determine which.
The day-to-day range may reflect genuine variation in need, visitors or periods away from the property, and should be interpreted alongside continence and hydration information held by the practitioner.
The recorded decrease of -13.33% between baseline and comparison windows is a descriptive comparison only and no reference range has been supplied.

What to explore

Are there known continence, hydration or medication factors that would explain variation in bathroom frequency?
Is bathroom use for washing or bathing expected to appear as longer episodes than those detected?
What was happening on 1 May and 20 May that might explain the extremes in recorded counts?

What the data cannot tell us Bathroom episodes are derived from movement and occupancy sensing and cannot identify the purpose of a visit. and Absence of a detected event does not prove that an activity did not occur.

Assessment finding

Movement and Room Use

Green Moderate confidence

What was detected

Room episodes averaged 102.5 per behavioural day with 38.73 room transitions per day and a daytime-to-overnight episode ratio of 11.53.
Detected occupancy per behavioural day was 85.35 minutes in the kitchen, 64.40 in the bedroom, 59.83 in the living room, 8.27 in the hallway and 7.12 in the bathroom.
The longest recorded daytime gap between consecutive room episodes was 178 minutes, and hourly episode counts peaked at 12:00 (301) and 18:00 (259).

Pattern identified

Movement is concentrated between 09:00 and 19:00 with twin midday and early-evening peaks, and very low episode counts between 03:00 and 08:00.

Why this may matter

The high transition rate may be consistent with independent mobility around the property, though telemetry cannot describe gait, steadiness or use of aids.
The 178-minute daytime gap may reflect a period of rest, seated activity in a sensor-sparse area, or time away from the property; the absence of exit telemetry means this cannot be resolved.
Total detected occupancy is well below 24 hours per day, so a substantial proportion of time is not represented by any room episode and should not be interpreted as absence.

What to explore

Are there areas of the property without sensor coverage where the individual regularly spends time?
Does the observed daytime peak pattern match what the individual or carers describe about their day?
Are the low 03:00-08:00 counts consistent with the individual's reported sleeping and rising habits?

What the data cannot tell us Room episode derivation depends on sensor placement, which has not been independently verified for continuity or like-for-like coverage. and Telemetry cannot describe quality of mobility, balance or falls risk.

Assessment finding

Overnight Activity and Quiet Periods

Amber Low confidence

What was detected

Across 20 analysed nights the average longest quiet period was 57 minutes, the median 40 minutes, the shortest 14 minutes and the longest recorded nightly value 197 minutes on 1 May.
On 18 of the analysed nights the longest quiet period was under 120 minutes or there were at least three bathroom episodes; no night recorded three or more overnight bathroom episodes.
Average first morning activity was 08:29 (509 minutes after midnight, standard deviation 110 minutes across 19 observed values) and average last evening activity was 23:29.

Pattern identified

Most nights contain frequent detected activity with short longest-quiet periods, alongside few overnight bathroom episodes.

Why this may matter

Short overnight quiet periods may reflect genuine overnight movement, a second person or pet in the property, or sensitive sensors registering minor motion; these cannot be distinguished from telemetry.
Overnight movement and quiet periods are not proof of sleep, waking, sleep quality or sleep disruption, and no sleep inference should be drawn from these figures.
Variability in first morning activity (standard deviation 110 minutes) may be consistent with a flexible daily routine and should be interpreted alongside any reported daytime fatigue.

What to explore

Does anyone else, or a pet, move around the property overnight?
How does the individual describe their own night-time rest and rising times?
Is the bedroom sensor positioned so that a person resting in bed would still register movement?

What the data cannot tell us The evidence can describe overnight activity and quiet periods but cannot confirm sleep, waking, sleep quality or restorative sleep., Only 20 nights were analysable within the requested period, and 19 assessment days contain no telemetry (warning coverage:missing-days)., and A plausibility check on overnight quiet periods was recorded as passed with 5 of 20 nights below a 30-minute short-quiet threshold (validation:overnight_quiet_periods).

Assessment finding

Routine Consistency

Amber Low confidence

What was detected

First-activity and last-activity consistency scores were both 0.0, with standard deviations of 293.4 and 577.75 minutes across 22 observed days.
Mean cosine similarity between each behavioural day and its next three behavioural days was 0.4921.
Daily room transitions ranged from 13 on 21 May to 87 on 15 May.

Pattern identified

The shape of the active day is moderately repeatable hour-by-hour, but the times of the first and last daily activity vary substantially.

Why this may matter

Low start and end consistency may reflect a genuinely flexible lifestyle, variable support visits, or days with gaps in telemetry capture at the edges of the coverage window.
The consistency score is a scaled derivation against a fixed 120-minute reference and does not carry any clinical or normative meaning.
These scores should be interpreted alongside known appointments, care visits and social commitments.

What to explore

Are there scheduled visits or appointments that would explain the more active days such as 15 May?
Does the individual describe their daily routine as regular or deliberately flexible?
Could partial-day telemetry at the start or end of coverage be distorting first and last activity times?

What the data cannot tell us Consistency scores are clamped derivations and no comparison population or reference range has been supplied. and The 22 behavioural days are not contiguous with the requested period, limiting trend interpretation.

Assessment finding

Property Exits and Community Access

Grey Not-reportable confidence

What was detected

No reportable property-absence metrics were available for this period.

Pattern identified

No pattern can be described because exit telemetry is not available for this domain.

Why this may matter

The absence of exit data means that periods with no in-home activity cannot be attributed to leaving the property and should not be interpreted as either social engagement or isolation.

What to explore

Would a confirmed front-door or exit sensor add value to future assessment?
What do the individual and their network report about frequency of leaving the home?

What the data cannot tell us The property exits plausibility check was skipped because the metric was unavailable (validation:property_exits_per_day). and Exit-related availability evidence was suppressed and cannot support any finding.

Potential anomalies

Patterns for closer review

On 1 May 2026, 34 bathroom episodes were recorded against a period average of 9.91 per behavioural day, alongside 158 movement episodes and the longest overnight quiet period of 197 minutes.

May reflect a day of unusually high in-home activity, visitors, or repeated passage past the bathroom sensor.
Could indicate a transient change in need, or a sensor behaving differently on that date.
Should be interpreted alongside any care records or reported events for that date.

Question for review Is anything known about 1 May 2026 — such as visitors, illness or a change in support — that would account for the marked increase in detected bathroom episodes?

On 21 May 2026, the final observed day, counts fell to 27 movement episodes, 13 room transitions, 13 kitchen episodes and 4 bathroom episodes.

May reflect a partial day of telemetry capture at the boundary of the coverage window rather than a behavioural change.
Could indicate time spent away from the property or in an unsensored area.
Should be interpreted alongside the note that telemetry ends on 21 May while the assessment was requested to 30 May.

Question for review Should the final observed day be excluded from trend interpretation as a partial-capture day, and is the reason telemetry ceased after 21 May known?

On 20 May 2026 only 1 bathroom episode was recorded, against a period average of 9.91 per behavioural day, while 58 movement episodes were still detected that day.

May reflect a period spent away from the property during the day, with in-home movement occurring at other times.
Could indicate a temporary bathroom sensor issue, since general movement was still detected.
Absence of a detected event does not prove the activity did not occur.

Question for review Was the bathroom sensor confirmed as functioning normally on 20 May 2026?

Review and limitations

Professional review

Questions for professional review

Does the kitchen-centred pattern of frequent short presences match what the individual and their support network describe about daily living?
Are there known health, continence or medication factors that would contextualise the bathroom frequency and its recorded decrease across the period?
How does the individual describe their night-time rest, given that telemetry can only describe activity and quiet periods?
Would confirmed front-door or exit telemetry, and resolution of the unknown-classification events, materially improve the value of future reports?
Should this report be repeated over a fully covered period before any comparison against these figures is relied upon?

Quality warnings

Average property exits per behavioural day could not be checked because the evidence metric is unavailable.
19 assessment days contain no telemetry.
49594 in-period telemetry events retain an unknown classification.

Data limitations

Passive telemetry describes detected events and behavioural patterns; it cannot establish diagnosis, subjective experience, intention or causation.
The evidence can describe overnight activity and quiet periods but cannot confirm sleep, waking, sleep quality or restorative sleep.
Sensor installation continuity and like-for-like coverage across comparison periods have not yet been independently verified.
Absence of a detected event does not prove that an activity did not occur.

This report is a decision-support summary of passive home telemetry for service user SU001. It does not diagnose, determine clinical risk, assess eligibility, recommend care changes or establish causation. All statements are limited to detected events within 22 behavioural days between 30 April and 21 May 2026, and traffic-light statuses indicate prominence for professional review rather than clinical severity. Findings must be interpreted by a qualified professional alongside direct observation, the individual's own account and other records.