Facility optimization in simple terms means improving equipment, building, and operational performance by optimizing resource use.
An optimized facility reduces energy consumption, maintenance costs, and downtime expenses and increases efficiencies. All go toward improving the bottom line through enhanced productivity, cost savings, and customer satisfaction. Building owners and facility managers want to fully leverage the promise of facility optimization to create value, whether it’s direct savings and efficiencies or indirect value generators.
Machine learning (ML) is foundational to facility optimization, transforming large amounts of data from IoT devices, sensors, building management systems (BMS), building occupiers, and more into actionable insights for decision-making and automated building management processes. The buildings themselves become truly adaptive, intelligent, and agile systems, working to improve the value for occupiers, and by extension owners.
Increasingly, facilities are outfitted with technologies connected with a (BMS) platform. BMS platforms can run building systems on conclusions drawn through artificial intelligence and machine learning. This use case details the value of machine learning for facility optimization and the benefits of using the AI & Analytics Engine to quickly and affordably build the machine learning solution.
"Buildings currently account for around 40% of global energy use, with AI technologies offering opportunity to reduce energy consumption through better automation, control, and reliability and improve the safety comfort and experience of building occupants."
- Rav Panchalingam & Ka C. Chan (2019) A state-of-the-art review on artificial intelligence for Smart Buildings, Intelligent Buildings International
Current Challenges for Facility Managers and Operators
Modern buildings are becoming more multi-functional and crowded, with a higher level of interplay between the different systems involved, making facility management more complex. Increasingly facility managers are adopting ‘smart’ building technologies to create better experiences for occupiers, reduce the operational and energy costs of facilities and optimize the management and maintenance of facilities.
Buildings are often outfitted with sensors and IoT technologies. At its core, IoT and sensors do not provide intelligence, however, they enable systems to be analyzed and communicate with each other. AI can overlay intelligence to BMS systems, providing predictive insight and automation, making facilities more interactive and responsive to their occupants, and their needs, and enabling greater operational and energy efficiencies.
The Machine Learning Solution for Facility Management
In a broader context, ML can be applied across a variety of areas within facility management and deployed to connect to existing cloud or data storage infrastructure including a BMS. ML solutions can be applied to utilize the available data for the following:
Forecasting
Analyse historical data and make accurate predictions about operational needs, for example forecasting energy consumption.
Anomaly Detection and Fault Prediction:
Fault detection and diagnostic operations to determine if there are issues with the equipment, settings, usage, or systems so preemptive action can be taken.
Automating Equipment Groups and Systems
Buildings have many different systems, like HVAC systems, security systems, and smart appliances running concurrently. The Engine can automate the use and settings of these systems in real-time to provide the optimal experience for the building occupier and to reduce energy wastage.
ML-Based Facility Optimization
The AI & Analytics Engine provides a simple-to-implement intelligence overlay to an existing BMS or can be used to develop a new smart building system from the ground up.
Data from building occupants, BMS, IoT devices, and external data like weather can be ingested into the Engine, prepared, and used to develop ML models that provide predictions and insights that are fed back into an intelligent BMS for use.
The Engine is an end-to-end ML and data science development and operation platform, built to reduce the complexity and time it takes to develop and deploy ML solutions.
The high-level architecture below represents a type of virtual BMS extension, that treats the on-site BMS as a sub-system. Any higher-order intelligence implemented in the cloud would override the on-site BMS.
This overlay of intelligence on top of the BMS will generate cost-savings and occupancy experience enhancements, through the optimized control of the sub-systems.
Benefits of the AI & Analytics Engine for Facility Optimization
The Engine provides a fast and configurable solution, that can be flexibly integrated with existing BMS infrastructure to assist to gain a competitive edge through optimizing facilities. Moreover, the Engine is built to be accessible, which means no need for teams of data scientists to develop and run the solution.
Green Energy and Water Efficiencies
- Enhanced facility management decision-making & reduced human error, leading to reduced wastage.
- Manage and automate energy usage and water usage to meet actual facility needs to reduce wasted resources.
Cost Saving
- Replacing preventative and reactive maintenance with cost-effective predictive maintenance models.
- Reduced costs from efficiencies in energy consumption through automated heating, lighting, and appliance management.
Occupancy Experience Optimization
- Improved occupancy experiences through automated, controlled, and use-adaptive environments.
- Improved safety and security planning and controls.
- The smooth day-to-day experience of site management elements such as the signing in or signing out of staff, floor occupancies, traffic, and logistical operations.
- Improve temperature, lighting, and humidity control providing more comfortable conditions to occupiers
Enhanced Safety Measures
- Climate control in facilities that require highly sensitive control measures.
- Improve fire safety, maintenance, and security systems through automation and anomaly detection.
.svg)