Gathering Real-Time Energy Intelligence from the Industrial Internet of Things

Real-Time Energy Intelligence from Industrial IOT
AvatarNovember 10, 2016

The Internet of Things (IoT) is rapidly evolving and being adopted across industries. The digital era of data-driven intelligence is creating new businesses, as well as transforming old ones into various segments. IoT is changing the way we live and interact with day-to-day activities.


The power industry has found many novel applications of the IoT, which is transforming energy delivery. IoT technology is increasingly being applied to real-time reporting, just-in-time decisions on energy usage, demand response management (DRM), and many artificial intelligence (AI) areas. There is a huge opportunity to utilize digital sensors on industrial equipment to gather insights. It’s easy to imagine intelligent environments in which buildings can generate power on site or on the grid through networked lighting. The industrial IoT era will create more shifts toward data-driven intelligence and a whole new way of using business intelligence (BI) derived from data collected by new smart energy devices.


We recently completed such a project for an energy client. The project involved performing real-time analytics on data gathered from IoT sensors. The data generated enables various types of DRM analysis. This capability allows the client to participate in a regional DRM opportunity, which involves monitoring impact sites and assets in order to deploy energy conservation measures. The solution allows analysis on asset utilization and cost savings for effective decision making. It will also open new revenue streams, including new construction and equipment procurement.


We designed and implemented a complete IoT framework, including:

  • A method of capturing data from IoT sensors, which is then filtered, processed, and pushed into the cloud
  • Another level of code for complex data processing in the cloud
  • A custom user interface (UI), including several microservices and a Representational State Transfer (REST-ful) application programming interface (API) to source data for various graphs and key performance indicators (KPIs)

The sensors are attached to various electrical units – such as chillers and heating, ventilation, and air conditioning (HVAC) – throughout a site’s buildings. The IoT sensors are connected to a central gateway called Java Application Control Engine (JACE). The JACE is an embedded system device used to collect data at a central location from disparate sensors. Such devices allow full building automation, in which data generated at all the sensors is sent to a cloud network via JACE. JACE acts as a filter, pre-processing the location of data. Data available in the cloud is for analysis, including historical and real time.


A set of microservices were written to pull data from the cloud at regular intervals and create a JavaScript Object Notation (JSON) file. The microservices run as web services, performing auto background refresh and republishing the JSON file at regular intervals. A custom dashboard was built using Google-designed Polymer technology. The Polymer dashboard uses JSON data as input and microservices to pull from cloud. The REST-ful web services are used to access data under the client server interface. The user experience (UX) dashboard entails graphs and KPIs refreshed with IoT data available from the cloud using REST-ful services. The solution is hosted on Cloud Foundry, which provides platform as a service (PaaS), all custom libraries, and Postgres database support.


We encountered the following project challenges:

  • Challenge 1: Handling semi-structured data generated by IoT devices in real time, whereby data is dynamic, complex, and requires transformation for final reporting.To overcome this challenge, a Java framework application was written to transform data from semi-structured to structured format for reporting. The Java program also handles problems such as missing columns, split large files consisting of 1000+ dynamic columns, and varying timestamp data resulting from IoT devices that are functional and non-functional at different times.
  • Challenge 2: Building microservices on Cloud Foundry to communicate between multiple apps and web services.To overcome this, we learned Cloud Foundry’s architecture and conducted detailed trials of how the system works. Then we built Java framework code that contained all microservices. Additional tutorials helped in learning and implementing new technology.
  • Challenge 3: Using new technologies like Google-designed Polymer UI with the existing client set-up.Polymer UI is a new and emerging technology. We had to learn and experiment. The content was collated by training groups and used as a program for the team to install, learn, and grow expertise.
  • Challenge 4: Building awareness of how to use IoT data and the business benefits it offers.We conducted multiple workshops, business meetings, and sprint demos with client stakeholders. The people involved possessed expertise and knowledge in diverse domains. We achieved buy-in on the final solution based on the business need and validation of data to be generated from IoT devices.

The customer received a concept-ready demo for large facilities implementation. There are cost savings and revenue improvements resulting from the IoT data gathered. The data is also useful in making decisions about usage, maintenance, scalability, and productivity.


The extensive focus on identifying opportunities bring a large benefits to various businesses – from a bottom-line focus on improving utilization and cost reduction to a top-line focus on improved and additional revenue. IoT implementation improves revenue growth, asset utilization, employee productivity, and operation efficiency.


If you’d like to learn more about how real-time IoT analytics can improve decision making, get in touch.