NLP For Digital Manufacturing

Name: Patti Vrobel
Affiliation:
Phone:
E-mail: patti.vrobel@autodesk.com
Website:
Knowledge Required: Natural Language processing (Ideally)
Language parsing
application development (Python, C++, etc)
web-service development (Python, JavaScript, Json, etc)
Motivation: The trend for user experience in the future is natural human interfaces. NPL is fairly mature as a technology and there are some reasonable open source libraries. This is an experiment into the difficulty and effectiveness of this approach for applications that have typically only had mouse and keyboard input for years.
Description: Experiments with interacting with CAD/Manufacturing products / services via Natural Language processing
Objectives: Project entails exploration into natural language processing as a means to interface to professional products. The project will involve the development of a smart interactive manufacturability bot (M-bot) or app similar to "Siri" on iphones. Interns will develop M-bot as the "machine collaborator" integrated into an Autodesk product to help a design engineer or manufacturing engineer interact with the product and make decisions or execute commands. Technical challenges will be (1) develop a language framework for the user to query M-bot's expertise and return responses that will inform decisions and (2) develop "interviewing" methods to extract knowledge from expert users as a method of training a machine learning system for M-bot. Project will be scoped to meet the needs of the student's time and interest.
Deliverables: software design
demonstration
final project report
Other comments: This project has some element of research to it, but ideally we would scope it in such a way that the students can learn more software architecture and programming skills.

   D. Kevin McGrath
   Last modified: Fri Oct 20 09:31:13 2017