Andi Assess

AI-generated draft

Independent living assessment report

Independent living review: five weeks of home sensor evidence

The clearest pattern is a reduction in detected daytime activity across the five weeks, concentrated in the bathroom and kitchen areas, while overnight activity and bedroom-area activity stayed at a similar level. This is a change worth discussing with the individual rather than evidence of any single cause.

Assessment at a glance

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

Amber Moderate confidence

Overall daily activity and movement around the home

Activity was recorded every day, but the daily level and the amount of moving between areas both reduced, mainly after the first week and then...

Amber Moderate confidence

Daytime compared with overnight activity

The overall reduction is specific to waking daytime hours; the night-time pattern held steady throughout. This matters because it points towar...

Amber Moderate confidence

Bathroom-area activity

This was the largest reduction of any monitored area, in both how often activity was detected and how much detected time it represented, and i...

Amber Moderate confidence

Kitchen-area activity

Kitchen-area activity remained present every day but reduced in both frequency and detected duration, again during daytime hours. Because kitc...

Green Moderate confidence

Bedroom-area activity

No clear change was identified in bedroom-area activity. This contrast is useful context: it suggests the reductions elsewhere are unlikely to...

Amber Moderate confidence

Overnight pattern and morning timing

The night-time picture is consistent throughout: activity is detected on every night, with repeated short periods of bathroom-area activity an...

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

Service-user reference: 2f11c4f9-11e9-4ab8-b270-019fce0d0dc6_20260908_155730
Assessment period: 1 August to 6 September 2026 (37 calendar days, Europe/London). Usable activity data was recorded on all 37 days, with 36 complete overnight windows (22:00-06:00) and 2 partial windows at the start and end of the period.
Generated: September 13, 2026

Executive summary

A general review of independent living sensor data for this individual, undertaken without a specific referral question, to identify any patterns or changes over the assessment period that a practitioner should discuss, monitor or explore further.

Detected daily activity fell from about 155 episodes a day over the first 12 days to about 105 a day over the final 12 days (about 32% lower). The direction was the same at every comparison window tested, although the size of the fall is smaller (about 13%) when the unusually busy days of 5-7 August are excluded.

Most of the reduction happened early: weekly averages were about 167 episodes a day in the week beginning 3 August, then about 118, 113, 111 and 101 a day in the four following weeks, so the largest step was between the first and second weeks with a gentler decline afterwards.

The reduction is a daytime one. Comparing 1-18 August with 19 August-6 September, daytime (06:00-22:00) activity fell in the bathroom area (about 34 to 20 episodes a day), the kitchen area (about 46 to 35) and the bedroom area (about 32 to 26), while activity detected between 22:00 and 06:00 was unchanged or slightly higher in all three areas.

Bathroom-area activity showed the largest reduction of any monitored area, from about 47 episodes and 176 minutes of detected activity a day in the first 12 days to about 23 episodes and 79 minutes a day in the last 12 days.

Kitchen-area activity fell from about 58 episodes and 248 minutes of detected activity a day in the first 12 days to about 38 episodes and 150 minutes a day in the last 12 days.

Overnight patterns were steady: bathroom-area activity was detected on every analysed night, averaging about six episodes a night with three or more on 35 of the 36 nights, and no clear change in this level was identified across the period.

The first detected activity of the day became earlier, averaging about 05:11 across the first 18 days and about 04:34 across the last 18 days, within an overall range of 04:00 to 07:26 and a period average of 04:52.

Changes over time

Movement between monitored areas fell alongside the episode counts, from about 111 changes of area a day in the first 12 days to about 68 a day in the last 12 days, averaging about 85 a day across the period.
Bedroom-area activity showed no clear change in either episodes or detected minutes (about 44 episodes and 159 minutes of detected activity a day on average): the direction differs depending on which comparison window is used and the differences are small, which contrasts with the bathroom and kitchen areas.
Indoor light levels and room temperatures also fell in every monitored room over the same five weeks, for example kitchen light readings averaging about 1,182 in the week beginning 3 August compared with about 694 in the week beginning 31 August, and kitchen temperature about 24.4C compared with about 21.7C; these environmental sensors reported on every day of every week.

Overall interpretation

Taken together, the evidence describes a person who remains active every day and whose night-time pattern is unchanged, but who is being detected less often during the day, particularly in the bathroom and kitchen areas, with most of that change appearing after the first week of August. The available evidence cannot determine why: a change in daytime routine, health or energy, more time in unmonitored parts of the home, more time out of the home, or the seasonal shift also visible in the light and temperature readings would all look similar here. The most important limitation is that activity sensors cover only the bathroom, bedroom and kitchen areas, there is no entrance or exit sensor and no activity sensor in the living room, so time spent out of the home or in unmonitored areas cannot be assessed, and continuous sensor operation has not been independently confirmed. The most useful next step is to check with the individual and anyone supporting them what changed in their daytime routine from the second week of August, alongside a conversation about whether frequent overnight bathroom-area activity is a concern for them.

Assessment finding

Overall daily activity and movement around the home

Amber Moderate confidence

What was detected

Detected activity averaged about 123 episodes a day across the period, falling from about 155 a day over the first 12 days to about 105 a day over the final 12 days, with the lowest day on 15 August (75 episodes) and the highest on 5 August (228).
Week by week the average was about 167, 118, 113, 111 and 101 episodes a day, so the largest single step occurred between the first and second weeks, with a gentler decline afterwards.
Movement between monitored areas followed the same pattern, falling from about 111 to about 68 changes of area a day between the first and last 12 days, and averaging about 85 a day across the period.
Excluding the three unusually busy days of 5-7 August, a like-for-like comparison of 8-18 August with 27 August-6 September still shows about 13% fewer episodes a day, so those early days inflate the size of the fall but do not create it.
Across the whole period, activity was most often detected in the mid-morning (10:00 and 11:00) and late afternoon (17:00), and in the hour from 22:00, which was the busiest single hour and produced detections in all three monitored areas on all 37 days; 23:00 and the small hours were the quietest.
The longest single daytime gap without detected activity during the period was 324 minutes (about five and a half hours).

Pattern identified

Activity was recorded every day, but the daily level and the amount of moving between areas both reduced, mainly after the first week and then more gradually. A change of this size across two independent measures may be consistent with a change in daytime routine, energy or occupation, and is worth understanding before it is treated either as ordinary variation or as a decline in function.

Why this may matter

Changes in weather, visitors, health, medication, mood or planned activity outside the home could all produce fewer detections indoors.
Fewer detections in monitored areas may also reflect more time in the living room or outside the home, neither of which is covered by activity sensors.
The size of the change depends on which weeks are compared, so it is best described as a step after the first week followed by a gradual drift.

What to explore

Did anything change for the individual in the second week of August, when the largest step in the data occurs?
Is the individual spending more time in the living room or out of the home than earlier in the period?

What the data cannot tell us Detected activity counts describe sensor detections, not what the individual was doing or how well they were doing it., Continuous sensor operation across the comparison periods has not been independently verified., and The evidence cannot establish any cause for the reduction.

Assessment finding

Daytime compared with overnight activity

Amber Moderate confidence

What was detected

Comparing 1-18 August with 19 August-6 September, daytime (06:00-22:00) episodes per day fell in the bathroom area (about 34 to 20), the kitchen area (about 46 to 35) and the bedroom area (about 32 to 26).
Over the same two periods, activity detected between 22:00 and 06:00 did not fall in any area: bathroom-area episodes were slightly higher (215 over 18 days compared with 238 over 19 days), as were bedroom-area and kitchen-area episodes.

Pattern identified

The overall reduction is specific to waking daytime hours; the night-time pattern held steady throughout. This matters because it points towards daytime routine, occupation or where the day is spent, rather than towards a reduction in movement across the whole 24 hours.

Why this may matter

A daytime-only change should be explored alongside daily structure, social contact, appointments and any recent change in mobility, pain or motivation.
Because there is no entrance sensor, time away from the property is one plausible explanation that this data cannot test.

What to explore

What does a typical weekday look like for the individual now, compared with early August?
Are there periods during the day when the individual is regularly out of the home or resting in an unmonitored room?

What the data cannot tell us Day-part comparisons group episodes by the calendar date on which they started and cannot distinguish a person resting from a person being absent. and Absence of a detected event does not prove that an activity did not occur.

Assessment finding

Bathroom-area activity

Amber Moderate confidence

What was detected

Bathroom-area activity averaged about 33 episodes and 120 minutes of detected activity per day with usable data across the period, and showed the largest reduction of any monitored area, from about 47 episodes and 176 minutes a day over the first 12 days to about 23 episodes and 79 minutes a day over the last 12 days.
The fall in bathroom-area detections is also visible in the underlying sensor data, where positive detections in that area reduced week by week while bedroom-area detections did not.

Pattern identified

This was the largest reduction of any monitored area, in both how often activity was detected and how much detected time it represented, and it occurred in daytime hours while overnight bathroom-area activity continued at the same level. Reduced daytime bathroom-area activity may be relevant to washing and self-care routines and is worth checking directly rather than assumed.

Why this may matter

Consider alongside any known change in washing or grooming routine, continence, skin care, or confidence and safety in the bathroom.
Shorter detected periods may reflect a changed routine, use of a different facility, or help from another person, none of which the sensors can distinguish.

What to explore

Has the individual's washing or personal care routine changed since early August, and is any support involved?
Are there any difficulties with access, comfort or safety in the bathroom during the day?

What the data cannot tell us Detected room-area activity is not the same as a completed personal care task or time spent in the room. and The evidence cannot determine whether personal care needs are being met.

Assessment finding

Kitchen-area activity

Amber Moderate confidence

What was detected

Kitchen-area activity averaged about 46 episodes and 188 minutes of detected activity per day with usable data, falling from about 58 episodes and 248 minutes a day over the first 12 days to about 38 episodes and 150 minutes a day over the last 12 days.

Pattern identified

Kitchen-area activity remained present every day but reduced in both frequency and detected duration, again during daytime hours. Because kitchen activity is often linked to meal preparation and drinks, a sustained reduction may be worth checking against eating and drinking patterns.

Why this may matter

Consider alongside appetite, meal delivery or shopping arrangements, and any move to simpler or pre-prepared food.
Appliance-use evidence is not available in this dataset, so kitchen sensor activity cannot be linked to cooking.

What to explore

How is the individual managing meals and drinks now compared with earlier in the summer?
Has anyone else taken on more of the food preparation?

What the data cannot tell us Kitchen-area detections do not establish that meals were prepared or eaten. and Appliance readings are present but usage sessions have not been derived, so appliance use cannot be reported.

Assessment finding

Bedroom-area activity

Green Moderate confidence

What was detected

Bedroom-area activity averaged about 44 episodes and 159 minutes of detected activity per day with usable data, and showed no clear change: comparisons across short, intermediate and broad windows disagree in direction for both measures, with only small differences between them, in contrast with the bathroom and kitchen areas.
Positive detections from the bedroom-area sensors held up week by week while bathroom and kitchen detections fell.

Pattern identified

No clear change was identified in bedroom-area activity. This contrast is useful context: it suggests the reductions elsewhere are unlikely to be explained by a general loss of sensor coverage, although sensor continuity has not been independently confirmed.

Why this may matter

Steady bedroom-area detections do not indicate how much of that time was spent resting, sleeping or doing other activities.

What to explore

Is the individual spending more of the day in the bedroom, and if so is that by choice or because of comfort, pain or fatigue?

What the data cannot tell us Detected bedroom-area activity cannot indicate sleep, rest quality or how time was used.

Assessment finding

Overnight pattern and morning timing

Amber Moderate confidence

What was detected

Across 36 analysed nights, bathroom-area activity was detected on every night, averaging about six episodes a night, with three or more on 35 of the 36 nights and no night without any, and no clear change in this level was identified across the period.
The longest nightly period without detected activity averaged about two hours (median 104 minutes), ranging from 54 to 234 minutes, and also showed no clear change, with the result differing depending on the comparison window used.
First activity of the morning averaged 04:52 across the period and became earlier over time, from about 05:11 across the first 18 days to about 04:34 across the last 18 days, with individual days ranging from 04:00 to 07:26.

Pattern identified

The night-time picture is consistent throughout: activity is detected on every night, with repeated short periods of bathroom-area activity and typically no more than about two hours at a time without any detected activity, and the day now tends to begin earlier. These patterns are steady rather than worsening, but frequent overnight activity together with earlier mornings may matter for daytime energy and should be considered alongside the daytime reduction described above.

Why this may matter

Overnight bathroom-area activity should be considered alongside fluid intake, continence, medication timing and any urinary symptoms reported by the individual.
Earlier first activity may reflect light levels, room temperature, habit or discomfort; the data cannot distinguish between these.
Night-time falls risk, lighting and a clear route to the bathroom may be worth checking given the frequency of overnight activity.

What to explore

Does the individual experience their nights as disturbed, and is overnight bathroom use something they would like help with?
Is the route to the bathroom well lit and safe to use in the early hours?

What the data cannot tell us Overnight activity and periods without detected activity are not evidence of sleep, waking or sleep quality., Two overnight windows at the start and end of the period were only partially covered., and Continuous sensor availability was not confirmed for any of the 38 overnight windows.

Assessment finding

Home environment and data quality context

Amber Moderate confidence

What was detected

Light levels and room temperatures fell in every monitored room across the period, for example kitchen light readings averaging about 1,182 in the week beginning 3 August compared with about 694 in the week beginning 31 August, and kitchen temperature about 24.4C compared with about 21.7C; bathroom, bedroom and living-room readings moved in the same direction, and these environmental sensors reported on every day of every week.
All three monitored areas, and the environmental sensors, recorded data on every day of the period, and bedroom-area detections held up week by week while bathroom and kitchen detections fell, which makes a general sensor failure an unlikely explanation for the reduction; continuous sensor operation has nonetheless not been independently confirmed.
Six device health flags were recorded in the five weeks: low-battery flags from the movable hub device on 20 and 28 August, and isolated device alarm flags in the kitchen (3 August), bathroom (10 August) and bedroom (28 August and 5 September). No tamper events were recorded, and these flags are worth checking with whoever maintains the installation.

Pattern identified

The home became cooler and darker over the same weeks in which detected activity fell. This is neutral seasonal context rather than an explanation, but it is a reminder that late-summer changes in daylight, warmth and outdoor opportunity can alter indoor routines.

Why this may matter

Temperatures in the low twenties Celsius falling towards autumn may prompt a conversation about heating costs and comfort as the season changes.
The device health flags are few and isolated, but confirming device condition with the installation provider would strengthen confidence in future comparisons.

What to explore

Is the individual comfortable and warm enough at home as the season changes, and are heating arrangements in place?
Can the installation provider confirm that all sensors, including the hub, were fully operational throughout August and early September?

What the data cannot tell us Environmental readings describe conditions near each sensor and cannot be attributed to any behaviour or decision by the individual. and Device status and battery flags indicate device conditions only and say nothing about the individual.

Assessment finding

Areas and activities that cannot be assessed

Grey Low confidence

What was detected

Activity sensors cover only the bathroom, bedroom and kitchen areas; the living room has environmental sensing only and there is no entrance or exit sensor, so property exits could not be checked and time spent out of the home or in unmonitored areas cannot be assessed.
Three appliance-classified data streams recorded 9,061 readings, but usage sessions have not been derived, so appliance use cannot be reported and must not be read as zero use.

Pattern identified

Because large parts of daily life are outside sensor coverage, reduced detections in the monitored areas cannot be separated from time spent in the living room or away from the property. This limits how far the daytime reduction can be interpreted.

Why this may matter

If understanding time out of the home matters to this assessment, entrance sensing or a short activity diary would fill the main gap.
Any conclusion about reduced activity should be checked against what the individual and those around them describe.

What to explore

Would the individual consent to additional sensing, or to a short record of daily activity, to clarify the daytime picture?

What the data cannot tell us No evidence is available for living-room use, entrance or exit activity, or appliance use. and Absence of a detected event does not prove that an activity did not occur.

Potential anomalies

Patterns for closer review

Three consecutive days at the start of the period, 5-7 August, recorded markedly more activity than the rest of the assessment (228, 200 and 197 episodes, against a period average of about 123 a day).

Visitors, a period of illness, unsettled nights or a change in routine could all raise detections for a few days.
These days raise the starting baseline and therefore make the later decline look larger than a like-for-like comparison suggests.

Question for review Is anything known about what was happening at home during 5-7 August?

21 August stands out as a single busier day within an otherwise lower later period, with 182 detected episodes compared with 94 the day before and 100 the day after.

A one-off event such as a visit, appointment or household task could account for an isolated busy day.
A single day does not change the overall direction of the period and should not be treated as a trend.

Question for review Was there a visit or particular event at the property on 21 August?

Review and limitations

Professional review

Questions for professional review

What changed in the individual's daytime routine, health, mood or support around the second week of August, when detected daytime activity stepped down?
How is the individual managing personal care and meals now compared with early August, and is anyone else assisting?
Does the individual find their nights disturbed, given that bathroom-area activity was detected on every analysed night and mornings are starting earlier?
Is more of the day being spent in the living room or away from the property, which this installation cannot detect?
Is the individual comfortable and warm enough at home as temperatures fall, and would additional sensing at the entrance be proportionate and acceptable if a clearer daytime picture is needed?

Quality warnings

Average property exits per behavioural day could not be checked because the evidence metric is unavailable.
3 appliance-classified source streams contain 9061 readings, but usage sessions were not derived because approved activity-state and sessionisation policies are not yet available. Do not interpret this as zero appliance use.

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.
Absence of a detected event does not prove that an activity did not occur.
The assessment contains 36 complete local 22:00–06:00 calendar windows and 2 partial boundary windows. Activity was observed in 36 complete windows. Continuous sensor availability was confirmed in 0 of 38 intersecting overnight windows.
Sensor installation continuity and like-for-like coverage across comparison periods have not yet been independently verified.

This AI-generated draft supports professional judgement and requires human review. It describes patterns in independent living sensor data only and does not assess clinical risk, diagnose, determine eligibility or recommend changes to care.