MAP is our Analytics Platform for delivering models and solutions at scale and at speed

Uses an Integrated Analytics Stack to hide complexity letting you build solutions quicker and easier

Key Benefits of the Analytics Platform

  • It’s an integrated analytics stack, making it quicker to build models and solutions. Hides a lot of the complexity

  • You have control. Dynamically change the underlying data structure of your application and add new raw data sources and analytics as your requirements change

  • Modular platform with ability to run multiple instances of each module. Modules include: Importer, Raw Data processor, ItemFactory, Calculator and Invalidator

  • Web client interface lets you configure and manage MAP. Everything powered by our API allowing you to configure and extract data

  • Data Blocks and Data Versioning reduce disk IOPS and deliver lightning fast calculations

  • Fully modular and uses MongoDB as datastore. Can scale to any size

Example of solutions built using MAP

Key features that make MAP unique as an Analytics Platform

Delivering an Integrated Analytics Stack for rapid development

MODULES

MAP is an Analytics Platform that includes 5 different modules that provide fully extensible functionality for importing, pre-processing, calculating, invalidating and creating Items. Therefore, multiple instances of each module can run on separate servers providing the ability to scale with your requriements.

Importers

Enables sub-second import from multiple data sources. Loads raw data into a temporary buffer. New Importers can be easily added for new data formats and types

Raw Data Pre-processor

Processes the raw data in the temporary buffer and stores it in the MongoDB database

Item Factories

Creates and updates the Item structure for each Entity. Item Factories can be called from the API or from the Web Client to automatically create new Entities or to update existing ones.

Invalidators

Sets the conditions under which an Item will be re-calculated. Invalidators can be time based (i.e. at a specific time), event based (i.e. when a new raw data point is added) or calcualtion based (i.e. when a calculation delivers a specific result.

Calculators

Extracts Raw Data for the Item from the database and carries out the specific calculations for each Item created in the Item Factories

CALCULATION QUEUE

MAP generates its own Calculation Queue at runtime and then maintains it as new Items are added. This queue processes all the invalidated Items that require re-calculating. In addition, the MAP Calculation Queue is totally scalable. So, there is no limit to the number of Items placed on the queue. Item properties can be invalidated by API call or from the web client. An invalidated Item will automatically invalidate every dependent Item.

The Calculation Queue is managed in the Analytics Platform by the Invalidator module and is a key feature in delivering the Integrated Analytics Stack.

DATA BLOCKS AND DATA VERSIONING

Data Blocks and Data Versioning are a key part of delivering MAP as a high speed, high volume Analytics Platform

Data Blocks

Rather than loading and persisting all data for an Item, data can be broken up into chunks called Blocks. Therefore, Blocks provide huge efficiencies in real-time processing and reduce traffic to/from the database.

Data Versioning

Data Blocks are complimented by the MAP concept of Data Versioning. All Item data in MAP is versioned, including Blocks. A version is simply a unique timestamp.So,it allows users to query for the relative age of data, specifically when it last changed, and for calculated Items when the last calculation started and completed. A client application can then tell if data has changed without having to load the data itself.

It is this versioning technique that allows MAP to efficiently detect when calculated items need recalculating (referred to as dirtying as calculation).

See our Blog entry for more information

DATA TYPES

Any Data Type can be created in MAP and a number are available as standard such as grids (block and vector), lookup and scalar and time series values. Fully extensible so users can create new Data Type specific to your IoT solution. This feature is key in ensuring that MAP is a high speed, high volume Analytics Platform suitable for any application.

ALERTS

The Analytics Platform supports configureable alerts for any Item, for Groups of Items or for the entire system. These can be system alerts or alerts used to generte e-mail, text or api alerts to the end user.

REPORTS

Scheduled e-mailing of management reports run from R or from outputs to SQl Server. HTML reports can also be generated withing MAP and used as a simple dynamic reporting service.

RESTful API

A fully featured RESTful Analytics API is available to configure, upload and download your application.

UNLIMITED GROUPING HIERARCHY

The data structure has an unlimited grouping hierarchy and one that dynamically change by adding or updating the appropriate Item Factory.

We got some of our inspiration for defining MAP as an Intergrated Analytics Stack from here