# Use cases

## Use Cases for the *Forecast App*

The *Forecast App* can be used in numerous practical scenarios where the goal is to predict future values based on historical measurement data. Below you will find typical use cases with a description of the objective, the recommended configuration, and possible extensions.

***

### 1. **Predict energy consumption**

**Goal:**\
Forecast future electricity, gas, or water consumption on an hourly or daily basis to optimize energy procurement and detect anomalies at an early stage.

**Example configuration:**

* **Asset:** Meter (e.g. electricity meter)
* **Target Attribute:** Energy consumption (difference value, not the meter reading)
* **Feature Attributes:** `hour_of_day`, `day_of_week`, outside temperature (optional)
* **Forecast Length:** 24 (for 24 hours)
* **Context Length:** 168 (one week as context)

**Extensions:**

* Visualization in the dashboard
* Automatic comparison with target values from the *Calculator*

***

### 2. **Regulate indoor temperature**

**Goal:**\
Prediction of room temperature for optimal control of heating, ventilation, or cooling.

**Example configuration:**

* **Asset:** Room climate sensor
* **Target Attribute:** Temperature
* **Feature Attributes:** `hour_of_day`, `day_of_week`, current window position or CO₂ values
* **Forecast Length:** 12 (for the next 12 time units)
* **Context Length:** 48–72

**Extensions:**

* Combination with rule engine to automate building control
* Alerts when planned threshold values are exceeded or not reached

***

### 3. **Monitor humidity**

**Goal:**\
Early detection of potential moisture problems by forecasting humidity in critical areas.

**Example configuration:**

* **Asset:** Sensor in technical room, archive, storage room, etc.
* **Target Attribute:** Humidity
* **Feature Attributes:** `hour_of_day`, temperature, air exchange rate (optional)
* **Forecast Length:** 6
* **Context Length:** 24

**Extensions:**

* Integration with alarm system
* Linkage with ventilation control

***

### 4. **Forecast utilization of meeting rooms**

**Goal:**\
Detection of recurring usage patterns for better room planning.

**Example configuration:**

* **Asset:** Occupancy sensor or occupancy status
* **Target Attribute:** Occupancy (0 or 1)
* **Feature Attributes:** `hour_of_day`, `day_of_week`
* **Forecast Length:** 48 (e.g. for the next two days)
* **Context Length:** 96

**Extensions:**

* Linkage with *Booking Widget*
* Display of free time slots on digital door signs

***

### 5. **Detect technical anomalies**

**Goal:**\
Indirect prediction of malfunctions, for example in ventilation, pumps, or servers - e.g. through rising temperature or changed power consumption.

**Example configuration:**

* **Asset:** Technical component with sensors
* **Target Attribute:** Operating temperature or power consumption
* **Feature Attributes:** `hour_of_day`, current load, outside temperature
* **Forecast Length:** 6
* **Context Length:** 72