Lessons Learned from a Smart Building Project

Experience is the best teacher, which is what we discovered as the lead for an Intel smart building project. Even though the lessons we learned may seem simple or obvious, they can be difficult to carry out or easily glossed over due to the hectic pace of a construction project. This was a Greenfield building; however, most of the following learnings also apply to Brownfield projects.



We outfitted a 10-story, 630,000 sq. ft. office building in Bangalore, India with approximately 9,000 sensors used to track and optimize temperature, lighting, energy consumption, and occupancy in the building. The implementation was based on Internet of Things (IoT) solutions designed to collect, analyze, and secure data from building systems, and increase the capabilities of the building management system (BMS).

Our office building is forecasted to use 40 percent less energy due to the implementation of smart features, according to our project lead Srini Khandavilli, IoT/Smart Building Program Director, Intel India.


A picture of a building.


Business goals

Organizations just starting a smart building project have many considerations, so it’s critical to establish clear goals. Even better is to identify the key challenges upfront. These were the main goals of our project.

  • Reduce resource usage. Improve the control of energy and water-related systems by adding features that are typically lacking in a static BMS.
  • Improve operational efficiency. Move to a mobile cubicle model to accommodate more employees in the building.
  • Increase occupant comfort. Eliminate oscillating temperatures in the building so employees aren’t too hot or too cold during the day.


At the beginning of the project, it’s important to clearly articulate what “success looks like” through measurable key performance indicators (KPIs). The metrics can help avoid mission creep by indicating when the project has met its business goals.


Use cases

When starting the project, we soon realized there‘s no single definition for a “smart” building, so we had to figure out which use cases to implement ourselves. We studied different ones and decided which best achieved our business goals:

  • Resource usage decreased when we added building analytics to control our energy sources: diesel generation, solar, fuel cells, and the grid. The solution also decreased our utility rates by enabling automated demand response.
  • Operational efficiency improved by installing occupant sensors in cubicles and tying them to a mobile cubicle booking application that helps employees find vacant cubicles, thereby increasing their utilization rates.
  • Employee comfort increased thanks to a machine learning algorithm that maintains a constant temperature in all building zones by taking more environmental factors into account.


Data points and API’s

We knew the reference architecture we were building would evolve and be the basis for multiple other implementations. We factored in data collection from a wide range of smart building touchpoints in order to maximize our ability to implement various building management applications, such as:

  • Energy monitoring and analytics
  • Smart grid demand response and demand-side management HVAC and lighting control
  • Remote asset monitoring
  • Operations and portfolio oversight

We also started insisting vendors expose their data over REST API’s. This enabled us to invoke and access these building subsystems from the cloud. The ideal state would be when all vendors embrace Open API’s, thereby enabling the extensibility of the underlying building subsystems.


IT partnership

It’s important to create a partnership with the IT department early in the project to ensure the smart building solutions are properly designed and interoperate with the company’s IT infrastructure.


Our results

The building analytics we added to reduce energy and water usage is forecasted to save $645,000 per year with a return on investment (ROI) payback period of less than four years. The mobile cubicle booking application increased the building’s employee capacity by approximately 30 percent. The machine learning algorithm that improves the building’s temperature control could improve worker satisfaction through increased thermal comfort by as much as 83 percent.1


Smart building opportunities

Today, the typical Intel office building uses a static BMS that may have limited capabilities to intelligently control energy and water-related systems. Making modifications can be very expensive due to the closed and proprietary nature of current BMS systems, which also is an impediment to incorporating the latest technologies and reaping the benefits of smart buildings.

At Intel, we’re moving to a BMS as a “Service” model that gives IT and facilities teams the flexibility to pick the vendors they deem best for specific functions. This approach also makes it easier to collect similar data from all buildings in the portfolio, enabling more intelligent decisions to be made at the portfolio level.

To learn more about the implementation of our smart building project, download the case study.

Published on Categories Embedded, Smart BuildingsTags , , , ,

About Srini Khandavilli

IoT Program Director, Intel Corporate Services Srini currently owns the worldwide strategy, roadmap and implementation of IoT SMART buildings at Intel. Srini has been at Intel for the last 18 years. His background includes managing large global software teams, software architecture, applications and middleware. He collaborates closely with the Intel IoTG group to ensure that the hardware and software assets deliver services meeting the needs of the SMART Building use cases at Intel.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.