OEM, SaaS Providers, Hardware OEMs

A Comprehensive IoT Software Platform for OEMs and SaaS Providers Shortens Time to Market and Reduces Software Development and Support Costs

SkySpark® is an extensible, end-to-end software platform for smart device data that can be applied from the edge to the cloud. SkySpark provides data collection, data storage, data management, automated analytics and data visualization for Internet of Things applications in the built environment and is designed to be used as a platform for OEMs to build their additional value on.

If you are planning to offer your customers solutions to collect, manage and create value from the data produced by smart, connected devices you will invariably encounter the need to buy or build software to address the following needs:

  • Connectors and APIs to access data from external devices, systems and applications
  • Examples: Bacnet IP, Modbus TCP, Obix, Haystack, SNMP, Sedona, OPC UA, MQTT, SQL, CSV import (manual & automated), and REST API
  • A Database optimized for efficiently managing huge volumes of “machine data” – time-series, multi-structured data with semantic modeling functionality
  • Historian functions and services optimized to query and process large volumes of timestamp value pairs
  • Analytics engine – to automatically process rules and algorithms against the data to detect issues, patterns, correlations, anomalies and trends in the data
  • Visualization – Ready-to-go Apps that automatically generate rich displays of data and analytic results, and tools that allow you to create customized user Apps and reports
  • A Distributed Computing architecture to enable deployment from the Edge to the Cloud
  • Multi-tenancy support – a single server can host multiple customers, each with their own secure data realm
  • Security: Certificate based authentication, TPM integration, HTTPS in the browser, LDAP support for centralized credential management
  • Automatic notification of detected issues via email with links and report attachments

SkySpark provides all of these functions in a ready-to-go platform that has been proven across thousands of applications.

SkySpark - A True Data-Focused Platform: Key Differentiators

If you are working with IoT devices and their data, you likely know that machine data is different than conventional IT data. SkySpark has been designed from the ground up for the unique needs of sensor, machine and device data.

Device data is by nature “time-series” data – a sensor value with a corresponding time stamp. High resolution device data can yield an immense volume of timestamp/value pairs. For example, if we store minutely data for a single sensor (we call a sensor a “point”) we end up with 525,600 samples per year. 100,000 points with two years of minutely data results in 105 billion samples! Although that isn't actually an extremely large system for SkySpark, a traditional database with 100 billion rows would be considered a really big database. In order to provide high speed processing of data and a responsive user interface for viewing and analyzing data, efficient storage and processing of these volumes of data is a top priority for any IoT software platform and an area where SkySpark excels:

  • SkySpark’s “Folio” database provides highly efficient data storage (>10 times more efficient than relational database technologies). A time-stamp value pair averages 12 bits of storage space in SkySpark. Most relational databases require 30-40 bytes to store the same data – some require much more. Beyond highly efficient storage of time series data, Folio adds the ability to model the meaning of data and define relationships between data items with semantic tagging. SkySpark uses the Project-Haystack.org meta data tagging methodology to represent the semantic meaning of data items.
  • SkySpark integrates a high-speed analytics processing engine with the Folio database. With SkySpark, analytic rules “run in the data” eliminating the delays and complexity found in software architectures where large data sets have to be pulled from the database and passed to a separate analytics process.
  • Analytic rules and algorithms are fully programmable to meet the unique needs of almost any data type, equipment system and applications – just add your domain knowledge to SkySpark’s library of over 500 analytic and data transformation functions.
  • System architecture is important too. SkySpark is scalable and distributable – it can run in everything from small “edge” devices to virtual servers in the cloud – on your favorite operating system
  • SkySpark supports both cloud or on-premise deployment for server level applications to address the needs of applications that simply can’t allow huge volumes of data to be transmitted to the cloud
  • This comprehensive platform provides lower Total Cost of Ownership than data technologies designed for business IT applications
  • SkySpark is proven and ready to deploy – today SkySpark is used across of 13,000 facilities consisting of over 1 Billion square feet of space (>90M square meters)

A Wide Range of Deployment Options to Fit Your Business Needs

SkySpark gives you the flexibility to deploy solutions that fit your customer’s needs and your business model – whether that is Software as a Service, on- premise deployment, or embedded products.

  • Install SkySpark as an end-use application for facility managers, energy managers. Install on a laptop of desktop or server, or offer as a cloud service (SaaS)
  • Use SkySpark as an in-house, tool for data acquisition and analytics - example for consultants involved in commissioning, energy analysis, monitoring and verification.
  • “White box” model – Use SkySpark as a backend database and analytics engine and integrate with other data visualization applications for final presentation of data and results to users
  • Embed SkySpark in small, low cost hardware devices (i.e., gateways, controllers) – collect data locally, process analytics at the edge, and serve rich visualizations directly from the edge device and/or replicate data to the cloud and serve visualizations from the cloud
  • SkySpark ‘s fully distributed architecture supports deployment across large numbers of nodes that work together in “clusters” providing a near-seamless user environment that enables users to navigate their applications from the Edge to the Cloud

Learn more about our OEM Programs >