SmartScan

During my final Electrical Engineering year, I worked on a very exciting project called Smart Scan. Initiated by a paramedical school in Eindhoven, Netherlands, the project aims to create a smart glove which can provide podiatrists with hard data and improve the orthosis making process.

The pitch video above was part of our submission for an engineering contest in which dozens of other projects participated. Smart Scan ended up receiving the second place prize.

How it works

Medical-grade positioning sensors are used to track the position and orientation of the technician’s fingers, and using software we have developed during this project, the data is filtered and a point cloud of the subject is created and displayed in a CAD program.

The video demonstrates the usage of the system and the current filtering capabilities.

My Contribution

My main contribution to this project was designing the software architecture and implementing the sampling and filtering pipeline. As this project was meant to span multiple semesters and teams, and was very likely to require improvements and changes along the way, modularity and scalability were key requirements.

Software Block Design Diagram

The software was implemented as a C++ library which could be used with different frontends (in our case a command line tool and a RhinoCAD plug-in were developed). Of course, an object oriented approach was taken and the sampling process was separated from the filtering stage. This allows for better filtering algorithms to be implemented lated without requiring changes to the rest of the pipeline.

Measurement pipeline

Results

The results produced at the end of this project were more than satisfactory, providing our client with a proof-of-concept prototype and a strong foundation for future work.

Comparison between Smart Scan(white) and an optical scan(green). The average error was found to be 2.9mm, with outliers as far as 25mm.

Publication

The work done during this project has also resulted in a research paper which was published in Sensors and is, as of writing this, pending review. The article can be viewed here.