What does it mean when your measurement is just within, or just outside of limits? Dealing with risks due to measurement errors and their implication on safety
{"title":"What does it mean when your measurement is just within, or just outside of limits? Dealing with risks due to measurement errors and their implication on safety","authors":"S. Mozar, W. Kao","doi":"10.1109/ISPCE.2017.7935017","DOIUrl":null,"url":null,"abstract":"In safety engineering there are occasions where a reading indicates that the result over the limit. For example if a measurement result is 5.51V, and the limit is 5.50V. Is such a reading outside the limit? Does a safety problem exist? The answer is not a simple yes or no. A number of considerations need to be made, for example how critical is this measurement for product safety? In other words is there a safety risk? To answer such a question we need to understand the effect of tolerances. This includes the tolerance of the measuring instrument used, and of the circuit being tested. First measurement errors are briefly reviewed, followed by a brief discussion of the statistical significance. Then statistical techniques and risk management are used to determine if the reading is out of tolerance and what is the likelihood of a safety issue. This may seem an “overkill”, but we consistently hear about consumer goods catching fire. One cell phone is banned by all airlines as the chances that it will catch fire is considered too high to put passengers' lives at risk. Safety engineering does not make adequate use of statistical analysis to understand and prevent unsafe conditions.","PeriodicalId":211888,"journal":{"name":"2017 IEEE Symposium on Product Compliance Engineering (ISPCE)","volume":"126 2 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 IEEE Symposium on Product Compliance Engineering (ISPCE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISPCE.2017.7935017","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
In safety engineering there are occasions where a reading indicates that the result over the limit. For example if a measurement result is 5.51V, and the limit is 5.50V. Is such a reading outside the limit? Does a safety problem exist? The answer is not a simple yes or no. A number of considerations need to be made, for example how critical is this measurement for product safety? In other words is there a safety risk? To answer such a question we need to understand the effect of tolerances. This includes the tolerance of the measuring instrument used, and of the circuit being tested. First measurement errors are briefly reviewed, followed by a brief discussion of the statistical significance. Then statistical techniques and risk management are used to determine if the reading is out of tolerance and what is the likelihood of a safety issue. This may seem an “overkill”, but we consistently hear about consumer goods catching fire. One cell phone is banned by all airlines as the chances that it will catch fire is considered too high to put passengers' lives at risk. Safety engineering does not make adequate use of statistical analysis to understand and prevent unsafe conditions.