AMI Testing - what does it take to get it right?
AMI components, like meters and endpoints, are in different locations in the field while using different protocols and hybrid communication technology (RF, PLC and GPRS) with the support of hardware like concentrators, repeaters and third party communication service providers. Head-end systems may also be scattered geographically while Meter Data Management servers are at a centralized location. AMI testing is done at the laboratory level and it requires the right automation and testing tools to test all components and scenarios across a wide range of geographies. This testing environment must be capable of replicating this complex environment for mass deployment testing. Some of the proven testing approaches are given below:
The meter is the primary component for data creation and it needs to be tested for accuracy of data capture while remaining in line with the specifications of the utility. Testing can be done by a meter testing laboratory that is certified by a regulatory board/authority. An in-depth understanding of different protocols and testing tools are also required at the laboratory level to evaluate the performance.
Endpoint plays the role of a gateway between the outside world and the meter. It is important to ensure accurate communication between meter to endpoint, endpoint to meter, endpoint to network and network to endpoint. Appropriate testing tools having communication capabilities with endpoints are required to verify the configuration parameters and compare them with the result report.
Communication Network testing:
Hybrid communication technologies are required for a complete AMI implementation for any utility. The Testing team should also have experts in different communication technologies. AMI network faces issues like time synchronization through various components, network capacity, speed, missing data packets and intervals. Network is one of the most complicated components to test in a laboratory. A Tester can test with few endpoints, concentrator with Radio frequency but mass level testing at laboratory level is difficult to achieve. Mass deployment testing can be done in a laboratory by using a simulator which has the capabilities to create networks with thousands of different components. Each component's behavior needs to be tested in actual condition and recorded in a simulator to replicate a near to real scenario behavior in laboratory. In actual field conditions there are possibilities of frequency jams and obstructions; which is not possible to replicate in a laboratory.
Head End System testing:
Utilities require single or multiple head end systems depending on the type of AMI components. Functional and performance testing need to be done for data collection, validation, configuration and messaging with different communication technologies and a number of components.
Meter Data Management Testing:
MDM support millions of data including interval meter data import, VEE and storage, daily meter determinant calculations and billing exports. MDM testing needs to be done to benchmark the level of performance and scalability. It can be done by performance tools with simulators that are needed to create millions of data.
End-to-end Scalability testing of AMI system:
End-to-end AMI systems work fine at the time of initial pilot projects with a small number of components; but the main roll out has thousands of meters, endpoints, other network components and large volumes of data collected on a daily basis. Performance testing should be done for end-to-end AMI systems to find out its scalability.
There are many tools available in market. It is also good to integrate the tools and reports of testing output, automation of testing process and simulators to generate End-to-end network components for scalability testing, which helps in effective AMI testing and to implement a robust system in field and serve long term to utility with the confirmation of ROI.