Mouldbox is applying research in machine learning to automate the quoting, design and manufacture of composite tooling. We pass these benefits on to the customer to reduce cost and improve lead times.
The best way to understand the Mouldbox technology is to see it for yourself. Our technology is deeply integrated into our website and powers our pricing, scheduling and DFM feedback. Try out the example part upload on our homepage to get an instant quote to design and manufacture tooling of a composite part.
The Mouldbox team brings together experts in machine learning and experts in composite tooling to build solutions to real engineering problems. Our tooling design is tightly integrated with our manufacturing processes which means we get real world feedback.
The revolution in machine learning offers many opportunities to improve the composite tooling manufacturing process. We chose to start the journey by automating the quoting of composite tooling. Mouldbox uses stochastic gradient descent and optimization to pick how to best make a mould.
Our next steps include the application of reinforcement learning methods in the design and manufacturing process, data mining of part designs and building in user feedback methods to educating our users in mould design.
For modelling problems with exact solutions geometry analysis is an excellent solution. Mouldbox technology makes research in geometry and topology analysis relevant to composite tooling. We are committed to deploying this technology and making it accessible on the web. Feature recognition, undercut analysis and visibility maps are all areas critical to the mould design process.
Our technology builds on many open source technologies. Examples of libraries integrated into the Mouldbox product are Tensorflow, Open3D and OpenCASCADE.