# US 2017/0286497 A1 · System for capture, analysis and storage of time series data from sensors with heterogeneous report interval profiles

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2017/0286497 A1
- Document ID: US20170286497A1
- Status: Published application
- Kind: A1
- Application number: 15/091,563
- Filed: 2016-04-05
- Published: 2017-10-05
- Earliest claimed: 2016-04-05
- Assignees: Fractal Industries, Inc.
- Applicants: Fractal Industries, Inc.
- Inventors: Jason Crabtree; Andrew Sellers
- Web page: https://patentgrep.com/patent/US20170286497A1

## CPC

- G06F 16/2477: {Temporal data queries}
- G06F 16/2264: {Multidimensional index structures}
- G06F 16/24568: {Data stream processing; Continuous queries}
- G06F 16/283: {Multi-dimensional databases or data warehouses, e.g. MOLAP or ROLAP}
- G06F 16/951: Indexing; Web crawling techniques
- G06F 16/953: Querying, e.g. by the use of web search engines
- H04L 67/12: specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A system and method for the capture, storage and analysis of multidimensional time series data from sources with heterogeneous reporting profiles has been devised. Data from sets of sensors that send variable amounts of multidimensional data at irregular intervals is received by a data processing device which processes the raw data to extract the parameters of interest and held until a preset number of sensor events or amount of stream data is received. The data is then committed to a data store until time of analysis. Older data is written to archive storage. The system allows complex selection and transformation of data store data by robust query language.

## Claims

### Claim 1

A system for capture, analysis and storage of data time series from sensors with heterogeneous report interval profiles comprising:
a data stream management engine stored in a memory of and operating on a processor of a computing device;
a multidimensional time series data store stored in a memory of and operating on a processor of a computing device;
a data query and output engine stored in a memory of and operating on a processor of one of more computing devices;
wherein, the data stream management engine:
(a) receives a plurality sensor data;
(b) aggregates the sensor data over a predetermined amount of time, a predetermined quantity of data or a predetermined number of events for transmission into the multidimensional time series data store;
(c) transmits a specific quantity of aggregated sensor data per each access connection predetermined to allow reliable transmission to and inclusion of the data into the multidimensional time series data store;
(d) transparently transmits quantities of aggregated sensor data too large to be reliably transferred by one access connection using a further plurality access connections to allow capture of all aggregated sensor data by the multidimensional time series data store under conditions of heavy sensor data influx; and
wherein the multidimensional time series data store:
(e) stores aggregated sensor data in a simple key-value pair with very little or no data transformation from how the aggregated sensor data is received;
(f) stores data for a predetermined number of samples then purges the oldest data; and
wherein the data query and output engine:
(g) is a point of interaction to set up analysis prior to sensor data collection by specifying such parameters as number of events or time units to be placed within each quantum of aggregated sensor data, the number of connection lanes between the data stream management engine and the multidimensional time series data store, number of lanes that can be combined at one time to transfer aggregated sensor data to the multidimensional data store, number of interrelated dimensions to be store per sensor;
(h) has an SQL like query language to retrieve sensor data of interest from the multidimensional time series database in a useful format;
(i) is the point of interaction for selection of transformations performed on the retrieved multidimensional time series data store as well as specifying the format of data output.

### Claim 2 (depends on 1)

2) The system of claim 1 wherein, a long term data store stored in a memory of and operating on a processor of a computing device receives older time series data that has timed on the multidimensional time series data store and stores that data until needed or released.

### Claim 3

3) A method for capture, analysis and storage of data time series from sensors with heterogeneous report interval profiles, the method comprising the steps of:
(a) receiving data from remote sensors with irregular reporting interval profiles;
(b) aggregating the sensor data based upon the number of sensors included in the analysis and a predetermined time interval or a predetermined number of events;
(c) transferring aggregated sensor data to a multidimensional time series data store using one to a predetermined plurality of communication lines dependent upon the number of sensors reporting data;
(d) storing un transformed aggregated sensor data in a key-value pair data store for a predetermined period of time, removing the stored aggregated sensor data as it times out;
(e) retrieving aggregated sensor data of interest to an ongoing analysis from the multidimensional data store using an SQL-like query language;
(f) transforming then formatting that data as best suited to its role in the ongoing analysis;

### Claim 4 (depends on 3)

4) The method of claim 3 wherein, aggregated sensor data that has aged to the point of being removed from the multidimensional time series data store is stored on an external storage server until needed or released.
