Some of our team members use 3D printers very often and have experienced with the frustrations of failed prints, and the 3D printers often have print failures including print coming off the bed, layer shifts, air bubbles, collapse of structure. Hands on experience with the frustrations of failed prints
What it does
In order to reduce waste of material and print time, we came up with the idea to use video recognition to quickly analyze the occurrence of failed prints and to notify the operator/owner as soon as possible through a SMS message.
How we built it
Used Windows 10 IOT to connect to raspberry pi for live streaming and continuous photo capturing. We also used an Arudino connected with CMUcam5 Pixy to detect if the print itself has moved changed in its x direction. The arudino would be connected to the raspberry pi and used to send a SMS message using the Twilio SMS API.
Challenges we ran into
- OpenCV compatibility
- Python to C# communication
- Many decision changes
- Working with new languages
Accomplishments that we're proud of
- Installing and connecting to Window 10 IOT on Raspberry Pi
- Running C# app on Raspberry Pi
- Completing the entire planned workflow
What we've learned
- The problems with trying new languages
- The benefits of having an IoT system.
What's next for 3DCam
- Integrating with Octoprint or PrintRun to have automated print pausing,
- Concept a design that will work with a variety of printers
- Tackle other reasons how a print can fail.