Startup Idea: Automated Environmental Data Cleaning System

Startup idea to develop a software that automatically cleans large environmental sensor data sets. This software would relieve the need for manual review by detecting and flagging errors or artifacts resulting from failing or fouled sensors.

I work in environmental science. As part of my work I collect large data sets using environmental sensors. The resulting data consists of a large number of time series records with different environmental variables (e.g. temperature, pressure etc). However, the data often contains errors or artifacts due to sensors failing or becoming fouled.

At present a technician has to go through all these records manually and flag the bad data. Ideally a software system would clean these records automatically.