Manual administration of meter readings is time-consuming, and often more than 10% of meter readings around the world are prone to human error. In the case of Blicker, the customer or field worker takes a photo of the meter, and Blicker’s technology applies AI to provide high-precision and accurate meter readings.
NMi’s rigorous testing and evaluation of applications like this are important for the energy sector. It helps build confidence in the accuracy of meter readings, which also ensures billing accuracy. NMi’s Technology Accuracy Evaluation was used to perform an objective evaluation test of the Blicker solution in meter-reading extraction.
NMi’s Head of Innovation, Mike van der Heijden, adds: “It’s imperative to validate new technologies in the sector. NMi’s mission is to support and provide a trust basis for new and existing measuring technology among developers/OEMs and end-users. To that end, NMi seeks out and partners with innovative technology start- and scale-ups like Blicker to develop new testing methodologies that future proof measuring accuracy in the utility sector as technology advances.”
NMi concluded that 99.79% of the given ‘first try’ inputs are deemed to have a high enough validity percentage to be usable in practice, saving utility companies the need to perform manual readings.
NMi evaluated 1000 pictures processed with Blicker and concluded that 869 could be classified with high confidence, 113 with medium confidence, and 18 with low confidence. In the case of the evaluated dataset, high confidence meant that there is a 99,79% chance that the read-out is valid. High and medium confidence read-outs together result in a 98,96% validity percentage.
Victor Westerwoudt, CTO of Blicker, says: “As our energy systems and energy use change significantly, our need grows for digitally connected and integrated solutions that are reliable and accurate. We are delighted to have worked with NMi, a trusted name in evaluation and testing. Their conclusions about the accuracy of our products provide an extra level of confidence in the outcomes of our reading technologies.”