1. The Purpose of AI Recognition

In order to allow the printing status to be monitored in real-time, AnkerMake has designed a built-in AI camera for its M5 printer. Thanks to the powerful Xburst chip, the M5 printer is capable of monitoring its printing status in real-time. The purpose is to detect printing errors in a timely fashion to reduce printing time and prevent potential loss of printing materials.

2. The Principle of AI Recognition

When the AnkerMake Slicer does the slicing, it generates images of simulated perspectives at different heights.

Additionally, when the AI is enabled, every time the M5 prints a layer, the camera will capture a photo and compare it with the images of simulated perspectives locally. Once the first layer print is done, the Z-axis will also raise to 50mm to help the camera to get a clearer visual comparison.

3. Common Identifiable Printing Phenomenon

• Spaghettification；
• Layer shifting on the Y-axis；
• Nothing is printing when the filament extrudes；
• Clogged nozzle/Filament leakage；
• Camera blocked.

1) M5 can identify the above printing phenomenons, but it currently cannot directly identify the exact error type;

2) Since the camera shoots from the right side of the X-axis, if the profile of the misaligned image on the X-axis is basically unchanged, AI cannot recognize the misalignment of the X-axis;

3) When the camera is blocked, in order to improve the accuracy of recognition and reduce the impact of potential wrong printing operations, it usually analyzes at least five consecutive layers before giving an alarm notification. This is the exact reason behind users being unable to receive a warning immediately when the camera is blocked.

4. Influencing Factors of AI Recognition

The AI algorithm works based on the brightness of the current ambient light, the color saturation of the wire, the color difference between the filament and the hot bed, and the accuracy of the image comparison.

In addition, users can turn on/off AI recognition, set up the sensitivity of AI detection/pause after alarm when an error is detected in the AnkerMake app.

5. AI Recognition FAQ

Q1. Why does my model print correctly but AI still triggers alarms?

This is a false alarm situation. AI alarms have many factors that affect them. Sometimes, false alarms may occur due to ambient light and color interference. When this happens, you can try to improve the lighting conditions, change the color of the filament, and adjust the sensitivity of AI to "low" to prevent triggering a potential false alarm.

Q2. Why is there an error in my printed model but no alarm?

AI recognition is judged by accumulating layer by layer. In order to identify phenomena more accurately, it usually needs to accumulate at least five layers, so there will be a certain lag. If this situation occurs, it is recommended to adjust the sensitivity of AI to "high" in order to be notified as soon as possible.

Q3. Can I still have AI recognition when using a third-party slicing software?

No, AnkerMake slicing software is required to generate AI comparison simulation images, so other third-party slicing software cannot support AI recognition function.

Q4. Can I still use AI recognition if my printer is not connected to the Internet?

Yes, AI recognition function is based on local computing. Thanks to the XBurst processor, AI recognition function can be used even if the printer is not connected to the Internet, but the AnkerMake app will be unable to receive any AI error prompt.

Q5. Will AI recognition be affected when printing at night?

Camera has a built-in infrared light. When the environment is dark, the infrared light will turn on to allow the printer to continue functioning even at night, but the printer’s AI recognition rate will be lower.

Q6. Will AI recognition upload photos/videos to the cloud?

The actions recognized by AI are processed locally, when an AI alarm occurs, a real-time image will be taken and sent to the AnkerMake app through the cloud for the user to confirm the error.