Even as World Community Grid computations continue, the team behind the Clean Energy Project is busy improving their database structure, making their results more accessible to the wider scientific community, and authoring papers to disseminate the methodology behind the project. Also, look out for their new project website coming soon!
The work Ed and Kewei have done on machine learning has been put into a paper which has just been submitted for review - we will keep you updated on how it progresses. Thanks to Ed and Kewei for putting this together. Ed and Miguel will present this work in a poster at the New England Machine Learning Day on May 18th, alongside the work they have done on sampling methods, which could improve the number of high-performance molecules that projects like the CEP produce.
Kai has been working hard on putting together a new CEP website, which - whilst not ready for release quite yet - should be something exciting in the near future! The website will feature much more searching capacity (something made possible by the new database structure we have put in) and even more information about each molecule.
One of our new undergraduates, Wendy, has just received funding to work in the lab this summer and will use this new database structure as part of her summer project. She will be working on a diverse range of problems, one of which will be to allow users to search for molecules that are similar to a molecule they know; this is something we think will really aid researchers looking for new materials for organic photovoltaic (OPV) devices.
She will also be working with another of our new undergraduates, Claudia, on new ways to visualize the CEP data. Ed has worked with some experts in 3D rendering at Harvard to produce new animations that explain what the CEP is about; we are now looking to extend this idea into providing a new way to interact with the database.
Sadly, Gregor has now finished his Masters studies with us, and left to go back to the ETH in Zurich. Gregor’s work on new calibration methods was very successful. Ed is now putting two papers together based upon this work, one on the method, and one on the dataset which it was derived from. We will keep you posted on how they progress. Thank you for all your hard work Gregor - come visit us soon!
As ever, we are blown away by the support that all of you have given our project. We are so grateful for your continued support - we simply would not be able to do it without you!
Keep Crunching!!
Your Harvard CEP Team