Some thoughts during mini-hackathon

1. Measurement should start automatically 1 min after sensor is in the drink. This is because values drift over time.
2. Batch effect/normalization for each day should be done (e.g. to set tap water as 0). This is because there are drifts over days, but relative positions in grid stay similar
3. Original measurements should be saved for reuse (e.g. before taking mean)