BTL Mark: Resolve interoperability issues & increase buyer confidence
|Intelligent Buildings with AI
Buildings may be constant structures, but the cascade of disruptions on automated building stakeholders is definitely evolving for the better to create efficiencies and new ways of living and managing the property.
In the world of
Technology Disruptions, buildings are one constant structure where we
live, work, shop or sleep in. When the building is a home, it comes
with amenities to create conveniences and make us feel safe. When the
building is a factory or office building, the focus is on efficiency
and keeping people who work there productive. When it is a retail
building, the focus shifts to the experience of the retail customer. In
hospitals and multi-tenant buildings, the complexity increases to serve
multiple stakeholders while still keeping the cost down.
How do all these different types of buildings become smart? Buildings and all the appliances and equipment inside are becoming connected with sensors and connectivity to the Internet. These sensors produce a large volume of data about the health of the building structure, safety of the environment, efficiency of energy usage and operational usage of the building assets by various stakeholders. Artificial Intelligence is needed to create intelligence from the time series of building data. This data becomes the training data to build AI models that make buildings smart.
What does a smart building do?
A smart home keeps us safe. A smart factory or commercial building should run efficiently, increase safety, and create conveniences for the occupants, facilities managers, building owners and all stakeholders.
So at a minimum, you can start looking at buildings to do the following:
Introducing Intelligent Buildings
An intelligent building with AI will do all of the above plus solve for problems that current buildings cannot do by adding convenience for facility managers, occupants, building supply makers and building owners.
Today, facility managers can use mobile apps from vacation to look at a dashboard of data about how their buildings are performing or click to do simple operations that need their authorization.
AI in the form of machine learning takes it a step further by making buildings intelligent. The following are some of the possibilities and how they can help stakeholder goals in the Automated Building value chain.
1. Energy management of buildings using Predictive Models
Appliance or equipment energy management machine learning models help reduce energy consumption. Figure 1 below shows the Machine Learning model for residential appliance energy consumption.
Learning helps offer a spatial and temporal model of urban
building energy consumption. See Figure 2 for the model developed by
the Student team at Stanford Urban Informatics Lab as seen in the
poster at Stanford Energy Week 2019.
Figure 2: Machine Learning model for Urban Building Energy Usage from
Stanford Urban Information Lab.
2. Sustainable energy management
Electric vehicle charging stations is common in buildings nowadays. Machine Learning (ML) models can predict the load on the utility grid to manage demand response. See Figure 3 for forecasting model developed by students using PECAN data set tracking EV usage in a residential community in Austin, Texas.
3: Forecasting model of EV Utilization from Public Austin City
PECAN data set.
3. Microgrids and sustainable
battery storage utilization
Companies are building machine learning models to take meter data and predict loads to manage the charge and discharge of batteries to save energy costs for buildings.
Enel X, the green tech unit of Italian utility Enel has taken it one step further to take smart meter data from EVs, Buildings, Solar and process this in real time to manage the energy storage and utilization for buildings.
See the Machine Learning models from Stanford Energy Week poster sessions and my interview with Rebecca Wolcoff, Machine Learning Engineer at Enel X in Silicon Valley.
Biometrics for Building Security
Security cameras watching building entrances are securing buildings by using machine learning to spot people and then applying facial recognition to identify people and tracking who is entering buildings. Fingerprinting is used commonly to identify people authorized to enter buildings. (See Figure 4)
4: Fingerprinting for
employee identification at
‘Letgo’ offices in Barcelona, Spain.
5. Robots and Drones
Robots come in many forms in smart buildings. Knightscope Robots watch the periphery of Buildings at Google and Juniper to augment the building security. They use AI, Anomaly detection to look out for abnormal patterns to alert the security team. See Figure 5 for the security robot.
Robots are perceived as a threat to people’s jobs and not as easy to adopt as Machine Learning models to look for patterns in data. Also, models not using publically available data and collecting videos and behavior data of building tenants bear risk from invading the privacy of users and the corresponding compliance to GDPR laws that govern them.
are used in buildings for aerial photography and surveying.
Recently drones seen flying near airports caused flights to be
cancelled at Heathrow airport creating security fears. So drones in
buildings are going to require more security checks before they can be
Buildings of all kinds will benefit from energy saving, operational efficiency and create new convenience for all stakeholders by adopting Artificial Intelligence. Artificial Intelligence today comes in the form of Machine Learning algorithms using public urban data, EV car charging data and private usage data of appliances and meters in buildings. AI in buildings is easier to adapt to using AI algorithms than physical robots and drones.
Buildings may be constant structures, but the cascade of disruptions on automated building stakeholders is definitely evolving for the better to create efficiencies and new ways of living and managing the property as we get to the connected Driverless World.
About the Author
Sudha Jamthe iis a Technology Futurist and the CEO of IoTDisruptions She teaches IoT and Autonomous Vehicle business courses at Stanford Continuing Studies and at DriverlessWorldSchool.com
[Click Banner To Learn More]
[Home Page] [The Automator] [About] [Subscribe ] [Contact Us]