Brightfield Real Time Energy Pipelines
Brightfield at the Google Cloud Leaders Circle Munich
Brightfield measures changes in industrial assets every second on Google Cloud. Like DNA sequences, slices of parallel IoT data allow Vertex AI to extract anomalies in power transformers, factory machinery and renewable energy at scale.
Above — real time magnetic fields captured 50 cm away through the air in real time from three phase cables. Heaters (pink), motors (turquoise). Brightfield hardware streams micro Tesla magnitudes at 50 Hz and 150 Hz (FFT at the edge in ARM embedded code) where they are then indexed (95th percentile) in Google Cloud Dataflow and Bigtable.
This accurately mirrors real power P, reactive power Q and power factor PF every second, valuable for re-dispatch energy markets.
Renewable energy production versus demand across thousands of transformers can be accurately predicted in real time.
At 50cm away, without proximity to dangerous 20 kV cables means fast installation. Legacy bulky grid equipment can take months to install; from the 450k transformers in Germany less than 0.1% are online and only deliver data every 15 minutes and not synchronised with each other (X axis). Brightfield securely installs in any transformer: getting it online fast and solving the issues of data synchronisation for energy assets at scale.
Brightfield also captures grid frequency divergence. Above — a large TSO gas turbine (peaks are about 90 minutes apart) compensating decentral wind energy. Balancing an increasingly decentral grid (divergence of 75 milli Hz) will be one of the most pressing issues of distributing future energy sources.
To scale an entire city with a team of 10 engineers (assuming 4000 transformers) takes about 1 year with Brightfield — rather than 20
Patent references here