NIOSHTIC-2 Publications Search
Potential problems associated with the rustrak ranger data logger's data storage technique.
Gressel-MG; Kovein-RJ; Hentz-PA
Am Ind Hyg Assoc J 1994 Oct; 55(10):970-976
An evaluation of the data storage function of the Rustrak Ranger (RR), a data logger designed to record and store industrial hygiene monitoring data was performed. The impetus for the study was the observation that data collected during a 2 hour monitoring study of methylene-chloride (75092) by a photoionization detector and stored in the RR lacked sufficient resolution and that the loss of resolution was due to the technique by which the RR stored data. The RR did not store data in a point by point fashion, but used a special storage method, known as adaptive storage, to maximize the available storage memory. Based on previously stored readings, the RR predicted the next reading to be either constant, linear, or exponential. If the reading fell outside a window surrounding the predicted value, it was stored. If it did not, it was discarded. The size of the window affected the rate at which the readings were stored and was adjusted, if necessary, so that data could be collected for the specified sampling time. If the windows were made too large, however, a loss in resolution of the data occurred. Previously recorded air monitoring data obtained for sampling times of 2 to 120 minutes were generated by a personal computer and sent to the RR. After recording the test data for each run, the RR was downloaded to another computer. The effect of sampling time on resolution of the data stored in the RR was evaluated by determining the average length of time that the stored values remained constant versus recording time and the percentage of data that remained constant versus recording time. The average lengths of time that the stored values remained constant and the percentage of data points that remained constant increased with increasing recording time. For recording times shorter than 30 minutes, the RR showed acceptable resolution by responding rapidly to variations in the signals. For exposure times longer than 30 minutes, it began widening the windows around the predictions of the incoming values, creating flat lines (constant responses) leading to a loss of resolution. The authors conclude that for industrial hygiene evaluations where exposures are expected to be highly variable, the RR should not be used to record data obtained with sampling times longer than 30 minutes.
NIOSH-Author; Equipment-reliability; Analytical-instruments; Workplace-monitoring; Computers; Data-processing; Statistical-analysis; Industrial-hygiene; Monitoring-systems
Issue of Publication
American Industrial Hygiene Association Journal
Page last reviewed: April 12, 2019
Content source: National Institute for Occupational Safety and Health Education and Information Division