Introduction
In this article we will walk through the different parts of the SensorX graphical interface that is used to configure the system and to quickly validate the icon and sound recognition.
In every menu and every view you will have a Info Panel on the right hand side which you can show and hide. In here you will find description texts, images and video sequences which explain the current selected or hovered configuration parameter.

Menus
Live Data
In the Live Data panel, you can see the live camera and microphone footages. If you have already trained icons or sounds, you will also find a graph with the detection signal.

By using the palette icon, you can change the colors of the different signals.

System
Open the system menu with the gear icon. Here you can find all system related information like the Serial Number, the Manufacturing Date, Hard-, Soft- and Client-Version. On top you are able to define an alias which can be displayed on the Hardware display.

With the Advanced Mode enabled you are able to see and adapt the following additional settings.
Note: If Advanced Mode gets disabled, the current settings will be kept.
Advanced CAN settings
It enables you to enable / disable single CAN data packets and to use user defined CAN IDs for each data packet.

Advanced Audio Detection
It enables you to use the parameters Min. Peak Distance and Min. Peak Prominence for teaching sounds.

Min. Peak Distance
It defines the minimum distance between the tallest peak to the others around it. All peaks within the configured distance to the talles peak will get ignored.
Min. Peak Prominence
It defines the minimum height of signal peaks to the surrounding signal. By increasing this value you are able to filter small peaks and noise.
Data Link 1️⃣
The Data Link settings are used to configure the Interfaces of the SensorX.
Ethernet
At the Ethernet settings you find the following parameters:

| Parameter | Possible Values | Description |
|---|---|---|
| MAC Address | (Hardware dependend) | Displays the unique hardware address of the SensorX Ethernet interface. The MAC address is fixed and cannot be changed. |
| IP Address | 192.168.88.24 | Defines the static IP address used to access SensorX via Ethernet. Make sure the IP address is compatible with the network of the connected device. |
| Subnetmask | 255.255.255.0 | Defines the subnet in which SensorX communicates. SensorX and the connected device must be configured within compatible network ranges to communicate directly. |
| Gateway | – | Defines the IP address of the network gateway. A gateway is typically only required when SensorX needs to communicate with devices outside its local network. |
| DHCP Client | ❌ / ✅ | Enable this option to let SensorX automatically obtain its network configuration from a DHCP server. When enabled, the network assigns an IP address to SensorX automatically. |
| Assigned IP Address | – | Displays the IP address currently assigned to SensorX by the DHCP server. This field is relevant when DHCP Client is enabled. |
CAN
At the CAN settings you find the following parameters:

| Parameter | Possible Values | Description |
|---|---|---|
| Termination CAN1 | ❌ / ✅ | Enable this option to activate the CAN bus termination for CAN 1. Use termination when SensorX is located at one end of the CAN bus. |
| Baud rate | 1000 500 | Defines the communication speed of CAN 1 in kBaud. Select a baud rate that matches the other devices connected to the CAN bus. |
| Prescaler | 1 | Defines the prescaler used to derive the CAN controller timing from its internal clock. Together with TSEG1, TSEG2, and SJW, this parameter determines the CAN bit timing. (Fixed Setting) |
| TSEG1 | 31 | Defines the length of the first CAN bit timing segment. Together with TSEG2 and the Prescaler, this parameter determines the position of the sampling point within a CAN bit. (Fixed Setting) |
| TSEG2 | 8 | Defines the length of the second CAN bit timing segment. Together with TSEG1 and the Prescaler, this parameter determines the CAN bit timing and sampling point. (Fixed Setting) |
| SJW | 8 | Defines the Synchronization Jump Width used by the CAN controller to compensate for timing differences between devices on the CAN bus. (Fixed Setting) |
Sensors 2️⃣
The Sensor settings allow you to configure each connected sensor individually. Currently, SensorX supports two types of sensors:
- Camera
- Microphone
As soon as you connect a sensor, SensorX automatically detects it and displays it under the corresponding port. The icon indicates which type of sensor is connected.

Camera
When a camera is connected, you can view its live feed. You can zoom in and out using your mouse wheel or the controls in the top-right corner.
In the Camera settings, you can configure the following options:


| Parameter | Possible Values | Description |
|---|---|---|
| Frame Rate | 10 20 25 50 100 | Defines the number of images captured by the camera per second. A higher frame rate provides a higher temporal resolution and allows SensorX to capture faster changes in the camera image. |
| Focus | 0 … 200 | Adjusts the focus of the connected camera. Change this value until the relevant objects or display elements appear sharp and clearly visible in the camera image. |
| Brightness | 0 … 1000 | Adjusts the brightness of the camera image. Choose a value that provides a well-exposed image. Avoid overexposure, which can cause colors to appear white and lose their characteristics, and underexposure, which can make color differences difficult to distinguish. |
Microphone
There are no settings available for the microphone. Make sure to position the microphone close to the speaker that produces the sounds you want to detect.


Detection 3️⃣
When you open the Detection menu for the first time, the setup wizard appears. ADMI guides you through the process of teaching the system to recognize objects, such as sounds and icons.
Right now it is possible to only use New video Recordings, later on it will be possible to use Video upload or Detection files from another configuration.

The wizard guides you through the following steps:
Sensor Recording
In the first step, select the sensors you want to use for object detection. In this example, the connected camera (Sensor 2) and microphone (Sensor 3) are selected.
Click Start to begin recording. You can then run your test case or trigger the icon, sound, or other event you want SensorX to detect.

We recommend using the Spacebar feature during the recording. Whenever an event occurs that you want to teach later, simply press the spacebar to set a trigger point. These trigger points are displayed during the teaching process, making it much easier to locate the relevant moments in the timeline.
Object Teaching
In the next step, select the types of objects you want to teach. Currently, SensorX supports teaching icons and sounds.
Additional object types, including tachometer readings, CAN IN signals, vibration patterns, and photodiode signals, will be supported in future versions.

Icon
When teaching icons, you can choose from different detection methods:

| Method | Description |
|---|---|
| Shape | SensorX automatically detects all edges within the defined search area and uses them to determine the shape of the icon. Based on this shape, SensorX determines whether the icon is active or inactive. During measurement, SensorX continuously tracks the detected edges and automatically compensates for camera vibrations and slight movements. This ensures stable and reliable icon detection, even if the camera position changes slightly. |
| Shape + Color | You can enhance the detected shape by adding color regions behind or around the icon. By combining shape and color information, SensorX can distinguish between different icon variants and states. This is particularly useful for indicators with multiple color states, such as Lane Keeping Assist indicators. Note: If the shape of an icon changes between different variants or states, SensorX cannot track them as a single icon. In this case, use the Brightness Spot detection method in combination with logic conditions. |
| Brightness Spot 🔒(Coming soon) | Define one or more small regions within the icon. SensorX analyzes the brightness of each region and converts it into a Boolean state based on the configured brightness threshold. This method is particularly useful for detecting simple On/Off states or illuminated indicators. By combining multiple Brightness Spots with the Logic Node Editor, you can also detect different icon variants, even if their shapes change. |
Teaching the Search Region
When teaching an icon by shape, the first step is to define the search area. Keep the search area as small as possible, as a smaller area allows SensorX to perform the detection faster.
You can define the search area using different frame options. In this example, we use the Free Form option. To position the search area precisely, zoom in or out using your mouse wheel or the controls in the top-right corner. Click and drag outside the icon to move the view.
Make sure that no other objects are visible within the search area, as they may interfere with the icon detection.

Teaching an icon by Shape
After defining the search region, you can enter a name for the icon and fine-tune the detection settings. First, navigate to a point in the timeline where the icon is active. If you set trigger points during the recording, as described earlier, you can use them to quickly find a suitable point.
Once you have selected a point where the icon is clearly visible, click Assign. SensorX analyzes the icon at this point and detects its shape.

You can use the controls on the right to adjust the color and intensity of the detected edges in the preview.
Next, click Verify to scan the entire timeline for the icon. Each period in which the icon is detected is marked with a blue bar. Review the results to make sure the icon is detected correctly.

If SensorX misses occurrences of the icon, increase the Detection Tolerance. If the icon is detected when it should not be, decrease the Detection Tolerance.
If adjusting the tolerance does not provide reliable results, select a different point in the timeline where the icon is clearly visible and click Assign again. You may also need to optimize the camera setup, for example by reducing reflections or avoiding changing lighting conditions.
Teaching an icon by Shape + Color
In the first step, teach the icon shape by following the same process described in the Shape Only section above.
Once the icon shape is detected reliably, proceed to the Color Regions step. SensorX automatically identifies suitable color regions within and around the detected icon shape.

By default, these regions are highlighted in yellow. You can adjust the visualization using the controls on the right-hand side.
Adjust the Identification Threshold until the relevant color regions are highlighted correctly. Use the Polarity setting to control which areas are identified as color regions.
In the next step, you can define different variants of your icon. Keep in mind that the shape of the icon must remain the same across all variants. Variants can only differ in the color of the detected color regions or in whether individual regions are active or inactive.

To define a new variant, click + Add Color Variant. Navigate to a point in the timeline where the new variant is clearly visible, then click Assign. The variant preview updates automatically based on the selected point.

Click Verify to analyze the entire recording. SensorX highlights all time spans in which the selected variant is detected, allowing you to check whether the variant is recognized reliably. Increase the Detection Tolerance if SensorX misses occurrences of the icon. Decrease it if the icon is detected when it should not be.

Once you have added all required variants and verified that they are detected reliably, click Finish Icon to complete the teaching process.
Sound
To teach a sound, zoom in on the relevant sound signal in the timeline using your mouse wheel or the zoom controls. Then click the waveform to select the Reference Region, indicated by the small gray area.

Once you have selected the Reference Region, SensorX calculates the FFT (Fast Fourier Transform) for the selected time span and displays the resulting frequency spectrum below.
For best results, we recommend placing the Reference Region shortly after the beginning of the sound signal, as shown in the example above.


Use the Loop function to continuously replay the currently displayed section of the recording. This can help you identify and select the relevant part of the sound signal.
In the next step, adjust the following parameters to define the key characteristics of your sound signal. For most sounds, the best results are achieved by selecting one or two distinct signal peaks and setting the Min. Peak Level relatively close to their peak levels.

Min. Peak Level
Defines the minimum threshold for detecting frequency peaks in the FFT spectrum.
For more reliable sound analysis, unwanted or irrelevant peaks can be manually removed from the table on the right.
Please validate your settings using the recording timeline to ensure the sound has been trained correctly.
Note: Depending on the environment and signal source, additional peaks may appear due to sound reflections, resonances, or other signal artifacts. These peaks can affect detection stability and accuracy.
Frequency Tolerance
Defines the allowable frequency deviation used during peak detection.
The parameter specifies a tolerance band around each learned frequency peak. During operation, detected signals are considered a match if their frequency lies within this band.
For each learned peak frequency, the system searches for corresponding signals within ± the configured tolerance value.
Cutoff Frequency
With the cuttoff frequency, you can define the frequency bandwidth in which you want to search for signals.
To improve reliability, it is recommended to keep the cutoff frequency range as narrow as possible and focus only on the relevant signal frequencies.
Please validate your settings using the recording timeline to ensure the sound has been trained correctly.
Note: Depending on the environment and signal source, additional peaks may occur due to sound reflections, resonances, or other signal artifacts. These peaks can reduce detection stability and accuracy.
Note: In rare cases, particularly when your sound signal has very low peaks, you may need to fine-tune the sound detection using the Advanced Settings. To access these options, enable Advanced Settings in the System Settings.
Sync 4️⃣
The time sync settings will be available in the near future.


Logic Editor 5️⃣
The Logic Editor allows you to use taught sounds, icons, and their variants and connect them to data outputs.
In future versions, you will also be able to combine multiple objects using logic conditions to create more complex detection scenarios.
| GUI Element | Name | Description |
|---|---|---|
| 1️⃣ | Auto create all | For a quick setup, use this button to automatically create input nodes for all taught sounds, icons, and their variants and connect them directly to the corresponding output nodes. The names assigned during the teaching process are applied automatically. |
| 2️⃣ | Clear all | Clears all nodes. |
| 3️⃣ | Icon | Adds an icon input node. |
| 4️⃣ | Sound | Adds a sound input node. |
| 5️⃣ | Output | Adds an output node. |
| 6️⃣ | Reset View | Resets the Node view for seeing all placed nodes. |
| 7️⃣ | Toggle Fullscreen | Activates Fullscreen. |
| 8️⃣ | Icon and Variant dropdown | Allows you to select the desired icon and one of its variants. You can choose either the icon shape itself or one of the defined color variants. |
| 9️⃣ | Name definition | Defines the channel name used for CAN and Ethernet data output. |

Data Output 6️⃣
General
| GUI Element | Name | Description |
|---|---|---|
| 1️⃣ | Sample Rate | Select the data sample rate. The following frequencys are available: 100, 200, 500, 1000 |
| 2️⃣ | Destination IP and Port | Select the destination IP address and port to which you want to send the data. |

Data Payload
In the Data Payload View, you can configure the structure of your data packets and output channels.
The first four data packets are reserved for system status information and are hidden by default. These Status data packets can only be disabled when Advanced Settings are enabled.
The Status data packets are followed by Userdefined 1 and Userdefined 2. These data packets can be configured to contain the output channels for your detected objects.

Use Auto Assign to automatically assign the output nodes defined in the Logic Editor to your User data packets. You can then drag and drop the assigned channels to change their order within each data packet.
You can delete data channels or lock them to prevent further changes. When a channel is locked, its name, description, and bit position within the data packet cannot be edited.
When a channel is unlocked, you can edit its name and description. You can also use the Assigned Output dropdown to assign a different unused output to the channel.
Expand the Status data packets to view the information they contain. For a detailed description of the available data, refer to the related SensorX Data Stream article.
Signal Out
Use the dropdown menus to select which taught objects you want to assign to the four available Signal Out channels. These signals are output as False (0 V) and True (5 V
electrically unaffected).

Launch 7️⃣
| Name | Description |
|---|---|
| Save Settings to SensorX | After changing any configuration parameters, make sure to save the updated settings to SensorX. |
| Discard Settings | Discards all changes that have not yet been saved to SensorX. |
| Set Default Settings | Resets all configuration parameters to factory defaults. |

| Name | Description |
|---|---|
| Autostart | Enables or disables Autostart Mode. When enabled, SensorX automatically enters Measurement Mode after startup. |
| Measurement Mode | This read-only field indicates whether SensorX is currently in Measurement Mode. |
| Stop Measurement Mode | Stops the Measurement Mode. |
| Start Measurement Mode | Starts the Measurement mode. |

Note: Configuration parameters cannot be changed while Measurement Mode is active.




