OVERVIEW
MAR-AM
MAR-AM is a web-based automated ML-assisted manufacturability analyzer and recommender for AM. This tool can serve as a first-level evaluation of designs for novice AM users such as designers to determine whether their given design is printable through their selected printer. As an ongoing project developed at McGill’s Additive Design & Manufacturing Laboratory (ADML), it has been a valuable research tool. By giving full access to the source, we hope this tool can be useful to lower the threshold of AM processes and support wider applications of AM in the future.
The main functions of MAR-AM can be summarized as the following:
- With the selected AM process, printer, process parameters, and material, the evaluation of manufacturability on the uploaded design will be provided before the real fabrication. It helps to avoid unnecessary printing failure and reduce the waste of cost.
- Specific recommendations on how to avoid printing failure will be given for each case based on the evaluation. Users can use those recommendations as the starting point to print their samples.
- Users are also welcomed to contribute to the development of MAR-AM by donating their printing data to the MAR-AM which can be used to update the manufacturability models.
AM data port
AM data port is a simple data port for data uploading and query specifically designed for ML in AM applications. This data port is web-based and shared with the public. Everyone is welcome to provide their open data or download and reuse the data. As this data port is mainly for machine learning (ML) applications, the attributes are categorized by keys of ML techniques. For each dataset, the donator needs to fill five required fields and five optional fields. The required fields include AM technique type, raw input data type, application/targets, whether the data is labeled or not, and the zip file for data. The optional fields include raw output data type, reference source, contact information, machine type, and material type.
Share your data
As introduced above, MAR-AM and AM data port will be continuously updated. You are welcome to share your data to support wider applications of AM.