A new report from Memoori Research says there is a young, but rapidly growing market for artificial intelligence (AI) and machine learning (ML) solutions in smart commercial buildings, which could be lucrative for an enterprising systems integrator focused on those markets.
Memoori defines smart commercial buildings as big box retail, specialty retail, malls, hotels, restaurants, office buildings, warehouses, data centers and mixed-use properties and other commercial real estate assets. Their recently released report does not include assessment of residential or industrial applications for AI technology.
While AI has seen usage in various forms in smart buildings for many years now, rapid innovation in domains including computer vision, natural language processing and especially generative AI promise to unleash new potential.
Memoori’s research identified 66 distinct use cases where AI is being actively developed or commercialized for use in the smart buildings market.
James McHale, Memori’s CEO, says the analysis aims to “parse hype from reality in this huge, complex, constantly evolving field shedding light on the steps that will be needed to be taken to fully unlock the AI’s potential.”
AI Growth Rapidly Rising
A meta-analysis conducted by Memoori for the report shows the broader AI market is predicted to reach over $500 billion globally by 2026, representing a 20.9% compound annual growth rate between 2023 and 2026.
AI front-runners are already seeing more than 15% faster annual revenue gains than peers less aggressive on AI adoption, the report notes.
Development of AI for smart buildings continues to accelerate, albeit from a small base at present. Of the broader AI market, commercial smart buildings accounted for only 0.5% of total revenues in 2020. By 2028, the smart building share of global AI could reach 0.8% according to estimates -- still small but significantly expanded on today's figure in relative terms, McHale says.
To date, most real-world deployments remain narrow in scope, but AI's immense promise has spurred over 300 identifiable companies to launch offerings into this space.
Memoori’s report notes that security and access control is “one of the most mature and established categories of AI use cases,” with the makers of major video surveillance and access-control systems already offering commercialized AI solutions.
Memoori says key applications leverage facial recognition, multi-factor biometric authentication, anomaly detection, intrusion alerts, behavior monitoring, automated visitor management, object detection, vehicle access control and crowd analytics.
For example, facial recognition for access control is commonly used, allowing authorized individuals to securely enter restricted areas without physical credentials. Video feeds are analyzed by algorithms that match faces to a database of approved personnel.
Biometric multi-factor access combines facial recognition with additional verification factors like fingerprints or iris scans for heightened security, with AI matching across biological characteristics.
Anomaly detection solutions establishes baselines of normal activity, continuously evaluating real-time data streams for deviations that may indicate security events or breaches, Memoori notes.
Better Emergency Response
There is also demand for AI and ML services in emergency and safety systems, focusing on the critical area of building safety. This domain explores the application of AI in enhancing emergency preparedness, detection and response.
Through continuous monitoring and analysis of data from various sensors and systems, AI facilitates real-time incident detection, optimized evacuation strategies, health and safety monitoring, and environmental risk management, the report concludes.
In real-time incident detection AI algorithms rapidly identify potential safety threats, such as unauthorized access or fire, enabling swift response to mitigate risks.
When leveraging real-time data for incident response, AI guides optimal evacuation and response strategies during emergencies to ensure occupant safety.
There is also critical importance of safeguarding interconnected smart building systems against cyber threats, Memoori notes. As buildings become smarter, integrating systems for lighting, HVAC and security, they also face increased risk from cyberattacks.
The use of AI and machine learning can help to provide a proactive defense mechanism, enhancing threat detection and response capabilities to protect against operational disruptions and ensure occupant safety.
Memoori points out that with network security monitoring, AI and ML can analyze network traffic and detect anomalies, providing real-time threat intelligence. AI-driven orchestration in cyber incident response allows for rapid response to security breaches, minimizing damage and downtime.
In IoT device management, utilizing AI for continuous monitoring and management of connected devices, mitigating vulnerabilities inherent in IoT ecosystems.
Opening More Doors
AI technologies hold immense promise for optimizing building operations and creating more sustainable, responsive environments, the report notes. But there are substantial near-term barriers slowing AI’s penetration across commercial real estate portfolios despite strong buy-in from C-suite decision-makers, the firm says.
Surveys reveal the top obstacles include a shortage of internal skills to deploy and maintain AI systems, integration challenges stemming from aging “legacy” infrastructure still prevalent in properties, and trust issues resulting from a lack of user confidence in AI recommendations.
Additional adoption headwinds encompass ethical questions around privacy and data transparency as well as looming regulatory risks currently taking shape differently across global jurisdictions.
Finally, even modern “smart” buildings often struggle with interoperability barriers and insufficient data quality, preventing AI from achieving results commensurate with its hype thus far, the report concludes.
Upgrading antiquated building equipment and disconnected monitoring platforms remains imperative before AI’s full potential can be unleashed holistically across lighting, HVAC, and security systems. Transitioning from proprietary technology “silos” to the adoption of open standards around data and APIs promises simpler integration, Memoori says.
From Narrow to Broad
McHale says proptech is still in a phase of “narrow” AI adoption where it’s being applied to specific use cases. However, computer vision -- and by extension, video analytics -- is more advanced and more “commercialized” than almost any other area of proptech Memoori analyzed.
“Where we are on the road from 'narrow AI' to AGI is not particularly clear, but the consensus of experts is that it’s not that far away,” he says.
McHale notes predictions that by the end of the decade AI will be smarter than human intelligence. “But smarter is an arbitrary term,” he says. “I take it to mean that it will be indistinguishable from humans in performing certain tasks that require intelligence, not that it will be self-aware or capable of extremely creative thinking.
“At the moment I think because of ChatGPT most people are thinking about AI like it's a tool, in the same way we thought about the Internet in 1996 as an email tool or a collection of Web pages,” McHale concludes.
“But AI is potentially much more pervasive. At some point all software becomes AI because it will write and improve itself. And we are already seeing the beginning of that -- possibly a layer of 'intelligence' that sits above all industries.”
