Odit-e propose machine learning algorithms able to process metering data to automatically digitize utilities’ networks.
SDG’s of application
Offers from michel clemence
Topology oofer is giving the connection scheme of final customers (transformer, LV feeder, phase, sequencing, trees…)
Low Voltage networks’ behaviors are largely unknown,therefore distribution network operators have to take big margins when planning their evolution. Odit-e has developed a way to process metering data in order to automatically build the network’s digital twin, from which network components can be extracted. Network simulation tools can therefore be automatically fed with appropriate network models, without requiring any manual effort: it then becomes possible for any utility to use such modern efficient tools This network identification has two main outputs: The network topology giving the connection scheme of final customers (transformer, LV feeder, phase, sequencing, trees…). This algorithm has already been validated Network characteristics (such as cable impedances or neutral grounding) can be extracted from the digital twin and sent to network simulation tools. This algorithm is still a prototype
Impact prediction processes metering data to automatically open the door to digital simulations.
Low Voltage networks’ behavior are largely unknown, so that distribution network operators have to take big margins when planning their evolution. The growing insertion of renewables, coupled with the need for flexibility, is pushing for a change. Distribution network planning need to evolve towards digitization. Odit-e has found a way to process metering data in order to automatically build the network’s digital twin, opening the door to digital simulations. It becomes possible to precisely estimate the impact of a new production, or electric vehicle charging station, to avoid unnecessary reinforcements and size the proper flexibilities. Network capacity maps can also be computed. By bridging physical and virtual worlds, Odit-e enable data driven decision making for flexible distribution networks.
Real time state estimation
An innovative way of providing real-time visualization for Low Voltage networks, by providing its most probable state.
Low Voltage networks have an extremely low transparency level. The current smart meters deployment is a huge opportunity, unfortunately they do not send information in real time. Up to day, real-time knowledge of low voltage networks stays out of reach, preventing any real-time management of flexibilities or predictive maintenance. Odit-e has developed an innovative way of providing real-time visualization for Low Voltage networks, by reconstructing its electrical “most probable” state from all the available information (primary substation, weather and calendar data). This statistical state estimation, made possible by combining electrical knowledge with artificial intelligence, drastically increase the transparency level for Low Voltage networks.