Performing the right measurement leads to making the right decision! Measurement uncertainty serves to indicate the quality of a measurement result. In order to fully utilize the results of tests, calibration or the corresponding data, it is necessary to understand and master all relevant associated uncertainties.
Proposed consulting services
To successfully carry out its missions of developing national references at the highest level, the LNE Laboratory deploys an extensive and versatile skill set in estimating measurement uncertainty. Thanks to this expertise, LNE is able to propose a menu of customized services to assist you with measurement uncertainty estimations in your particular context based on a range of benchmarks GUM*, GUM-S1, ISO 5725.
* Guide for expressing measurement uncertainty (JCGM: 100 or NF ISO 98-3)
From measurement process analysis to uncertainty calculation
How to go about calculating or evaluating measurement uncertainty?
The approach proposed by LNE in 4 steps
- Definition of the measurand, measurement process analysis, list of potentially influential factors, and model building
- Evaluation of standard uncertainties for each of these factors
- Propagation of the variances and covariances of individual factors
- Expression of the result with an extended uncertainty calculation and appropriate rounding of the measurement result.
This service requires becoming familiar with your site, analyzing and understanding your measurement process, and jointly crafting alongside you a detailed measurement model or perhaps a global statistical model.
Let's note that this method is also applicable to computation or numerical codes.
The set of tools implemented
- Process synopses
- 5M graphic models (or Ishikawa diagrams)
- Uncertainty calculation software
- Results sheet
- Flash training (link to the Intranet page) either customized or detailed (link to ME13), prior to the technical guidance service
- Creation and execution of an experimental campaign
- Partner-based R&D: Development of a method dedicated to uncertainty and adapted to your needs
- Statistical analysis of test and calibration results, plus compliance evaluation
- Statistical modeling of sensor behavior
- Signal and image processing
Why entrust LNE?
- Participation in producing benchmarks (GUM) and alternative methods (Monte Carlo or ISO 5725)
- COFRAC*-accredited experts and laboratories (ISO 17025) across all industrial measurement and testing fields
- Dual statistical and metrological competence
- Contribution to various European-scale research projects in both Applied Mathematics and Statistics
- Trusted third party.
* List of sites and scope of accreditation available on www.cofrac.fr