Use cases and examples
To help you understand the value of Flows in real-world scenarios here are some practical examples. Each of these use cases demonstrates how Flows can automate processes (business or consumer), improve decision-making, and create tangible user value.
Machine Overheating Alert
Problem:
As a maintenance engineer, I need to be alerted immediately if a critical machine overheats so I can prevent costly breakdowns or safety hazards.
Solution:
Continuously monitor the machine’s temperature sensor. If the temperature rises above 80 °C, automatically trigger an alert to the maintenance team via push notification and email.
Flow:
- Asset State as a trigger node - listens for incoming temperature data.
- Constant as a state node - a threshold value.
- Compare as a condition node - checks if the value is greater than
80°C. - Notify as an action node - sends an email and push notifications to the maintenance team.
Display Average Temperature
Problem:
As a tenant, I want to see the average temperature of my apartment on the dashboard to understand overall comfort.
Solution:
Collect readings from all sensors, calculate the average, and show it on a home display.
Flow:
- Asset State as a trigger node - listens for incoming temperature data.
- Average as a condition node - calculate the average.
- Publish New State as an action node - publishes
averaged temperature.
Help needed in Retail centre
Problem:
In a large retail centre for technical equipment there are several sections where customers may need help about a certain item or an assembly. However, it’s hard to know who from the personnel is needed when there is a general buzzer request.
Solution:
There are LoRa button devices and Radio gateway device deployed in a ground. When LoRa button is pressed at the electronic department, a specific tailored message is transmitted via radio to a support specialist. When they get to the customer they can acknowledge and reset the flow.
Flow:
Sustained Overheating (≥30 min) Alert
Problem
As a maintenance engineer, I only want an alert if a machine stays above the critical temperature continuously for 30 minutes, not for brief spikes. Short-lived peaks can be caused by normal factors and they don’t always require intervention, let’s say that short-live peaks are normal. What matters for me is the persistence: there the risks escalate. The goal is to reduce the false alarms and focus attention on incidents that are statistically more likely to indicate a true failure mode or unsafe operating condition.
Solution
Start a “high-temp” timer the moment the temperature first crosses 80 °C. The timer represents the start of a potential incident and runs only while the temperature stays above the threshold.
- If it remains > 80 °C for 30 minutes, send an alert (push + email) with device ID, site, start time, elapsed duration, and recommended first actions (e.g., reduce load, inspect cooling flow).
- If it drops to a safe level (< 75 °C hysteresis) before 30 minutes elapse, cancel the pending alert and reset the timer to avoid nuisance notifications from brief spikes.
Flow
Temperature, humidity and moisture control in the greenhouse
Problem:
A plant farmer wants to have the perfect conditions for his plants in the greenhouse. He has all the necessary equipment to do it but would need to constantly monitor and turn on the pumps, heaters and humidifiers himself. He needs to monitor the room temperature (in °C), humidity and soil moisture and adjust the conditions based on the calculations.
Solution:
Measure the temperature, humidity and moisture levels. Based on the comparisons with desired high and low level thresholds (constants) send commands (downlink messages) to control and adjust the cooler, heater and water pump units.
Flow:
This flow consists of 3 parts, each of them controls temperature, humidity and moisture levels respectively (these CAN be separated into 3 different flows):
Temperature control:
Humidity control:
Soil moisture control:
Smart farming: any eggs in there?!
Problem:
An animal farmer builds his business by providing only fresh organic eggs. Thus, he spends a lot of time walking around his farm checking if the chickens had laid any fresh eggs, because picking them only in the evening is not good enough for the business. He’d like to know exactly when a fresh egg has been nested.
Solution:
Set a weight sensor in the chicken nest. If the weight changes, the chicken left the coop, but if the weight is still different from the calibrated one (when the egg laying spot is empty) then there is an egg there.
Flow:
Smart Lighting Control
Problem:
As an office manager, I want lights to turn on only when people are present during working hours and turn off after 10 minutes of inactivity to save energy.
Solution:
Use a motion sensor and a timer to control lights. Each time motion is detected, lights turn on and the inactivity timer restarts. If no motion is detected for 10 minutes, the timer fires and turns the lights off.
Flow:
- Asset State (motion sensor) as a trigger node - listens for motion events.
- Send Command as an action node - turns lights on when motion is detected.
- Timer (Duration: 10 min) as a connector - restarts on every motion event; fires after 10 minutes of inactivity.
- Send Command as an action node - turns lights off when the timer completes.
Track Maximum Temperature of the Day
Problem:
As a building manager, I want to know the highest temperature recorded each day to monitor HVAC performance.
Solution:
Compare each incoming temperature reading against the stored daily maximum. If the new reading is higher, update the maximum. Reset the stored value at midnight.
Flow:
- Asset State (temperature sensor) as a trigger node - listens for new readings.
- Asset State (daily max, virtual asset) as a data node - provides the current stored maximum.
- Compare as a condition node - checks if the new reading exceeds the stored maximum.
- Publish New State (daily max) as an action node - updates the stored maximum when a new high is reached.
- Scheduler (midnight) as a trigger node - resets the daily maximum at the end of each day.
- Constant (reset value) + Publish New State as action nodes - write the reset value to the daily max asset.
Smart Cleaning
Problem:
As an office manager, I want to schedule cleaning only if a room was occupied for more than one hour per working day.
Solution:
Use a motion sensor and a timer to measure cumulative occupancy time. Pause the timer when the room is empty and resume it when people return. When the timer reaches one hour, schedule a cleaning. Reset the timer at midnight.
Flow:
- Asset State (Occupancy,
true) → Timer: Start / Resume - Asset State (Occupancy,
falsevia Invert) → Timer: Pause - Timer (Duration: 60 min) — accumulates occupied time across the day.
- Timer: Out → Increments the Counter
- Compare: Compares counter increment with Constant = 1, in order to order cleaning only once a day
- Webhook: Sends request for cleaning to cleaning service
- Scheduler (midnight) → Timer: Reset AND → Counter: Reset
