On the 3/15 sensor box deployment we noticed an issue where valve switching only started like 10 minutes or so following startup. It's unclear why this happens but it is unideal; we don't want to have to cut out bad data at the beginning of each deployment, and we don't want to extend the time it takes to validate and deploy each box because of a potentially preventable delay. We should recreate this failure, determine why its happening, and see how we can mitigate it (or I guess document that it happens).
On the 3/15 sensor box deployment we noticed an issue where valve switching only started like 10 minutes or so following startup. It's unclear why this happens but it is unideal; we don't want to have to cut out bad data at the beginning of each deployment, and we don't want to extend the time it takes to validate and deploy each box because of a potentially preventable delay. We should recreate this failure, determine why its happening, and see how we can mitigate it (or I guess document that it happens).