Editing Biochemical oxygen demand (section)

===Software sensor===
There have been proposals for automation to make rapid prediction of BOD so it could be used for on-line process monitoring and control. For example, the use of a computerised [[Self-organizing map|machine learning]] method to make rapid inferences about BOD using easy to measure water quality parameters. Ones such as flow rate, chemical oxygen demand, ammonia, nitrogen, pH and suspended solids can be obtained directly and reliably using on-line hardware sensors. In a test of this idea, measurements of these values along with BOD which had been made over three years was used to train and test a model for prediction. The technique could allow for some missing data.  It indicated that this approach was possible but needed sufficient historic data to be available.<ref name="Rustum-etal2008">{{cite journal |last1=Rustum |first1=Rabee |last2=Adeloye |first2=Adebayo J. |last3=Scholz |first3=Miklas |title=Applying Kohonen Self-Organizing Map as a Software Sensor to Predict Biochemical Oxygen Demand |journal=Water Environment Research |date=2008 |volume=80 |issue=1 |pages=32–40 |doi=10.2175/106143007X184500 |jstor=23804289 |pmid=18254396 |bibcode=2008WaEnR..80...32R |s2cid=24738186 |url=https://www.jstor.org/stable/23804289 |access-date=3 September 2021|url-access=subscription }}</ref>