Some thoughts I have so far:
1. Whatever the project is, it has to be something we'd be able to quickly iterate on (build-test-evaluate)
2. If there are types of sensors that are just now getting increasingly cheaper, we could probably use them (e.g. IMUs got cheaper when smartphones became popular). Same goes for types of data (e.g. high-res geospatial/satellite imagery).
3. High-impact projects revolving around energy, clean water access, food production, resource utilization and other humanity-level problems would be awesome. However, making something just for the sake of cool tech would work too.
Please feel free to share any thoughts/ideas you might have. Thank you!