How stillness detection works
Stillness detection works on two sides. The mobile app analyses the device's sensors, and the server independently measures time, and only when both agree is the full alarm triggered. As a result, a single sensor fault or a momentary loss of connectivity does not by itself set off an alarm.
Response times, confirmation times and sensitivity thresholds are configured in the panel. This means the threshold is not a fixed value written into the code, but a decision of the utility, which can be changed after analysing the first weeks of work.
The starting pattern and why it is only a starting point
As a starting point we know the warning pattern from an LTE watch: 20 seconds without movement gives a warning, a further 30 seconds without a response triggers the alarm, and the SOS button raises the alarm and calls the phone at once. In total, that is 50 seconds from the moment of stillness to the alarm.
For a worker walking along a corridor, this pattern is reasonable. For someone standing in a well and reading a pressure gauge, it may be too short. That is why good practice is a separate profile for confined spaces and a separate one for ordinary rounds, rather than one threshold for everyone.
Starting pattern of stillness times
What the data from a working deployment say
In a working stillness detection deployment, over 3,400 hours of monitoring were recorded, the median alarm closure was 52 seconds, and 7 out of 10 stops are cancelled by the employee alone. This environment is different from a well, but the conclusion is useful: most alarms end with a cancellation by a person, so the warning before the alarm works as a question, not as a notice of misfortune.
If after deployment in wells the share of cancellations is clearly lower than 7 in 10, this does not mean better vigilance. It may mean that the threshold is too long, or that people do not react to the warning because they do not know what it means.
How to choose thresholds step by step
Start from the starting pattern for rounds and introduce a separate, longer profile for confined spaces. Treat the next two weeks as an observation period: collect false alarms and cancelled alarms, and note the activities during which they occur.
Then shorten or lengthen the threshold by one value at a time, so that you know what had an effect. Changes are recorded in the log with their author and time, so they can be accounted for later.
Well and tank: a longer profile
Work in a well is often static. Lengthen the warning time, but keep a short response time after the warning, so that the decision about the alarm does not drag on indefinitely.
Make sure the worker knows how to confirm that everything is fine. The lack of a clear confirmation action is the most common reason for cancellations that come too late.
Ground round: a shorter profile
During a round, the 20 and 30 second pattern is a good starting point. Here the worker moves, so stillness has a clear meaning.
What happens after the threshold is crossed
When the threshold is crossed, the system sends a push notification and an email to the people responsible, creates an alarm permit from the ALARM template, marks the alarm on the map and can trigger an external webhook. When the employee moves again, the alarm permit closes on its own.
SOS is a separate track: it is active and unconditional, does not depend on the stillness settings and does not close itself when movement is detected. In a well it is worth telling people clearly: stillness is an automatic alarm, SOS is your decision.
Mistakes to avoid
The most common mistakes are one threshold for all work, no training on confirmation, and changing several parameters at once. Add to that the lack of a person who receives the alarm: technically everything works, but nobody reads the message.
Remember also that connectivity in a well is sometimes limited. For sites without coverage, we describe in a separate page the variant with an active card and a LoRa bridge.
Checklist
Video
Frequently asked questions
What stillness threshold should be set in a well?
Start from the 20 and 30 second pattern as a reference, then lengthen the profile for confined spaces after an observation period. The utility sets the specific value in the panel.
How do we know that the alarm is real?
The full alarm is triggered only when the app and the server, which measure stillness independently, agree.
Does the alarm close itself?
The stillness alarm permit closes itself when the employee moves again. SOS lasts until a person closes it.
Who changes the thresholds?
A person with permissions in the panel. Response, confirmation and sensitivity thresholds are configured by the utility.
Let's match the thresholds to your wells
We will plan with you the observation period and the threshold profiles for work in confined spaces.
Let's talk