Back to Basics is an article series that highlights important, but possibly overlooked, information facilities management professionals should know.
In facilities management, edge artificial intelligence (AI) can be transformative. Automating routine tasks and improving the accuracy of building data could streamline workflows and lower operational expenses. However, successful deployment requires management to understand what edge AI is and navigate common implementation challenges.
The Role of Edge AI in Facilities Management
Edge AI changes how facilities handle building data. Instead of relying on distant cloud servers, this technology processes information right where it is collected, enabling faster decisions and more responsive building systems.
What Is Edge AI?
Edge AI is a machine learning model deployed at the network periphery. It combines computing and AI technology to process and analyze information where it is collected. These algorithms run on Internet of Things (IoT) sensors and similar hardware, filtering information locally rather than sending it to a centralized cloud server.
How Does It Work?
Internet-connected sensors and cameras continuously collect building information. An AI model then processes and analyzes this information. Unlike standard AI, which relies on distant data centers, the edge model runs locally on nearby hardware. This setup allows for extremely low latency. Edge AI processes data within milliseconds, enabling real-time decisions without relying on the cloud.
What Does It Do?
Facilities managers can use edge AI in the same ways standard models can be used. Sensor information helps algorithms predict equipment issues before failures occur, reducing downtime and repair costs. Energy management systems analyze consumption patterns to adjust HVAC, lighting, and other systems for optimal efficiency throughout the day.
Edge AI improves the occupant experience by monitoring space utilization and environmental conditions. It automatically adjusts temperature, lighting, and air quality based on real-time occupancy and preferences. This technology can even analyze incoming maintenance tickets and building conditions to prioritize maintenance requests.
Edge models can run in the background to handle administrative tasks. Traditional property appraisal is only 70-75% accurate and can take up to three weeks, while AI-powered property valuation is 90-95% accurate and takes mere minutes, shortening the typical time frame.
The Infrastructure Supporting Edge AI
Machine learning models can only be self-updating and largely autonomous if they have the proper infrastructure. This includes hardware, software, and networking at the network’s edge. To ensure compute and connectivity, facilities managers must invest in technologies like storage, IoT devices, and low-latency networks.
To ensure the AI remains accurate and functions as intended, they should consider developing guardrails and governance frameworks. Orchestration platforms and security frameworks will also help ensure business continuity and system security. The specific preintegration steps they take will vary depending on where they build or outsource their ege AI system.
Key Implementation Considerations
While implementing edge AI is beneficial, facilities managers face several deployment challenges. Addressing these issues early ensures successful implementation.
Team-Wide Coordination
Due to physical obstructions and unreliable cellular networks, considerable communication challenges are common in noisy, crowded, and architecturally complex spaces. Edge AI addresses this by processing information locally, minimizing the need to constantly transmit large amounts of data. This makes it resilient to poor network conditions and physical obstructions that interfere with connectivity.
Increased Energy Usage
Data centers have existed for decades, but they need new, energy-intensive equipment to run AI workloads. While existing facilities often have a capacity of around 10 to 25 megawatts, hyperscale versions can reach 100 megawatts or more, consuming as much electricity annually as 100,000 households. Edge AI servers must be positioned close to the systems they monitor, which drives up electricity costs.
Notably, electricity prices will rise regardless of whether facilities implement edge AI. However, there are token and subscription costs to consider. Professionals should conduct a cost-benefit analysis and find ways to offset emerging expenses.
AI Model Maintenance
Edge AI models need to be efficient due to hardware and power constraints. Although machine learning models are self-updating, they may experience concept drift. If patterns in the information shift, accuracy drops.
For instance, if a building system’s logs switch from month-day-year to day-month-year format, the algorithm’s accuracy may degrade. This issue can also occur as resident expectations evolve or when the seasons change.
Despite operating under hardware and power constraints, edge AI systems require efficient processing. Quantization techniques and model compression can improve efficiency without sacrificing accuracy. This way, models remain responsive while meeting the strict performance and energy requirements of edge environments.
How Do Facilities Managers Benefit?
In facilities management, edge AI provides real operational benefits. From efficiency gains to lower operating expenses, the technology addresses core, everyday challenges.
Improving Efficiency
Edge AI identifies inefficiencies and reduces waste, continuously improving building efficiency. One research group developed a self-updating system to enable around-the-clock decision-making in rooms and open areas based on zonal management and user behavior patterns. This pilot platform reduced energy consumption and carbon emissions by over 25% by issuing real-time commands to terminal devices.
Gaining Real-Time Insights
Facilities managers and building occupants both benefit from immediate insights. Edge AI delivers this information without cloud computing delays, offering superior speeds and lower response times than other technologies, including standard AI.
Reducing Costs
These improvements lower operational, maintenance, and energy expenses. Edge AI streamlines tasks without human intervention and maintains occupant satisfaction. This reduces utility and labor expenses while helping secure contract renewals.
Supporting Business Continuity
Equipment can continue to function and perform AI tasks even when completely disconnected from the internet. Operations continue uninterrupted during network failures. In locations with limited or downed network infrastructure, edge AI hardware can analyze sensor information for safety, equipment health, or environmental changes, delivering critical insights locally rather than waiting for sporadic uploads.
Prepare Your Properties for Edge AI
Implementing edge AI into facilities management can solve common facility problems, such as unexplained high energy use or frequent maintenance delays. Understanding this technology, its functions, and infrastructure needs helps facilities managers make informed decisions about timelines, budgets, and vendors.
April Miller is a senior writer at ReHack. She has more than five years of experience writing on AI in business. You can explore more of her work on ReHack, Careers In Governement, ASBN, or connect with her on LinkedIn.