System and method for distributed, secure, power grid data collection, consensual voting analysis, and situational awareness and anomaly detection
A system and method for data collection and aggregation using a distributed network of communications enabled sensors connected to another primary network to achieve a secondary out-of-band monitoring perspective, for example, in power grids. The data collection system includes an aggregation and processing server configured to collect data from a variety of sensors adjacent to the monitored network each sensor includes secondary power such that it can continue data transmission even during power grid outages. The data collection system includes a method for secure real-time data ingest, machine learning enabled analysis, risk assessment, and anomaly detection on a broad geographic scale irrespective of isolated network boundaries.
1. A method for establishing power grid situational awareness to detect and interpret anomalies, the method comprising:
collecting sensor readings representing metrics of a power grid at least partially by (i) sending requests for said sensor readings from an aggregation server to a plurality of sensors of communication network nodes via one or more secure data collection channels and (ii) receiving said sensor readings at said aggregation server from said communication network nodes via said one or more secure data collection channels in response to said requests for said sensor readings, said communication network nodes (a) being configured to provide communication services to one or more locations that receive power from said power grid and (b) being on one or more communication paths that are separate from said power grid;
generating visualizations of said sensor readings; and
applying machine learning processing to improve visualization of sensor data and anomaly detection.
2. The method of claim 1 , wherein said machine learning processing incorporates power flow and threat modeling based on historical sensor data to increase predictive accuracy.
3. The method of claim 1 , wherein said machine learning processing incorporates classification error collection data to increase predictive accuracy.
4. The method of claim 1 , wherein one or more additional geospatial datasets are combined with said collected sensor readings to provide enhanced situational awareness across a plurality of dimensions.
5. The method of claim 4 , wherein said one or more additional geospatial datasets comprise one or more of utility data, weather data, and parcel data.
6. The method of claim 1 , wherein generating visualizations of said sensor readings comprises obtaining real-time data from a data stack.
7. The method of claim 1 , wherein visualizations of said sensor readings comprise one or more of sensor mapping, voltage over time, and power outage information.
8. The method of claim 1 , wherein applying machine learning processing to improve visualization of sensor data and anomaly detection comprises anticipating power outages through historical trends.
9. The method of claim 8 , wherein said historical trends include correlation between weather events and power outages.
10. The method of claim 1 , wherein said plurality of sensors is distributed across respective service areas of two or more different electricity providers.
11. The method of claim 1 , further comprising improving classification processing through classification error correction.
12. A method for establishing power grid situational awareness, the method comprising:
collecting sensor readings representing one or more metrics of a power grid using at least partially by (i) sending requests for said sensor readings from an aggregation server to a plurality of sensors via one or more secure data collection channels and (ii) receiving said sensor readings at said aggregation server from said plurality of sensors via said one or more secure data collection channels in response to said requests for said sensor readings, said plurality of sensors being electrically coupled to said power grid, and said plurality of sensors being associated with one or more communication network nodes (a) being configured to provide communication services to one or more locations that receive power from said power grid and (b) being on one or more communication paths that are separate from said power grid;
storing said sensing readings as a network data layer;
aggregating said network data layer with one or more secondary data layers to generate a stack of data layers; and
producing a geospatial visualization of said stack of data layers.
13. The method of claim 12 , wherein said plurality of sensors is distributed across respective service areas of two or more different electricity providers.
14. The method of claim 12 , wherein said one or more secondary data layers comprise one or more of a utility data layer, a weather data layer, and parcel data layer.
15. The method of claim 12 , further comprising using a machine learning application to improve quality of data used in said geospatial visualization.
16. A method for establishing power grid situational awareness, the method comprising:
collecting sensor readings representing one or more metrics of a power grid at least partially by (i) sending requests for said sensor readings from an aggregation server to a plurality of sensors via one or more secure data collection channels and (ii) receiving said sensor readings at said aggregation server from said plurality of sensors via said one or more secure data collection channels in response to said requests for said sensor readings, said plurality of sensors being electrically coupled to said power grid, and said plurality of sensors being associated with one or more communication network nodes (a) being configured to provide communication services to one or more locations that receive power from said power grid and (b) being on one or more communication paths that are separate from said power grid;
storing said sensing readings as a network data layer; and
applying machine learning to at least said network data layer to generate a threat model of said power grid.
17. The method of claim 16 , wherein said plurality of sensors is distributed across respective service areas of two or more different electricity providers.