Ask HN: How can I help as a software engineer in the current Covid-19 epidemic? As a person with software engineering skills, I want to help out in any way I can automate communication, perform number crunching or help towards building a vaccine (I am fairly certain there are some elements of software engineering involved) |
Help with efforts to develop at home testing kits: https://twitter.com/JMRothberg/status/1236748928177131525
Help with data visualization: https://twitter.com/jb_mouret/status/1236202203251200003
This virus is dangerous because it overwhelms hospitals with a lot of patients (death toll in Wuhan 6%, rest of China less than 1%).
Software can't magically fix things (I don't expect it to be a popular opinion here ;) especially if you don't have the required knowledge of the field.
Good luck.
1. Mine social media for possible cases.
Bluedot does not use social media, because it is very noisy, but a combination of human oversight and user flagging could potentially be useful. Reporting these potential cases to the CDC and to that user could be very useful. Example: https://www.reddit.com/r/nyc/comments/fayko1/my_covid19_stor...
2. Find ways to target ads at possible superspreaders or at-risk people for accurate information and advice. Example: I saw someone at the airport wearing cloth masks- those don't help. And worse, it confers a false sense of security.
3. Build a smartwatch app that buzzes when you try to touch your face.
If everyone tomorrow stopped touching their face in public, the virus would be greatly reduced.
4. Cough sound analysis for pneumonia detection
As we move from containment to mitigation, and the prevalence of COVID-19 surpasses that of influenza, detecting pneumonia earlier will enable lower mortality rates through earlier intervention.
5. A centralized source that finds all COVID-19 and SARS-CoV-2 related papers so that doctors, drug discovery researchers, etc. can easily find the newest and most promising research without having to manually search medrix, biorxiv, lancet, etc. This will likely be most useful for physicians to look at emerging treatment plans from China and South Korea.
6. Homomorphic encryption for GPS based path crossing and high threat location flagging. Two apps for this already exist in Korea, however in the US privacy is a greater concern. Homomorphic encryption might help with this.
7. Applying a discovery engine like AGATHA (https://www.reddit.com/r/MachineLearning/comments/f3fpih/r_a...) on the virus literature including new papers and making it usable to the public.
In general- think about small ideas that can be easily implemented (quickly!) and that don't try to solve the problem entirely but rather help in small bits.
Also, not software, but if any of you can build a compact and cheap far-uvc lamps (eg using a shg crystal and a blue diode laser or led like Sharp did), you could do a ton of work: https://www.nature.com/articles/s41598-018-21058-w. Alternatively, a microwave source capable of ~7-10 GHz should also work: https://www.nature.com/articles/srep18030