Honeywell's new AI tools target the false alarm problem
Honeywell launched EBI One, Distinguish detectors and Niagara 5 at Accelerate 2026, targeting false alarms and a shift toward autonomous building operations.

Why it matters
- Honeywell launched EBI One and new fire safety AI tools at its Accelerate 2026 event on 12 November 2026.
- EBI One builds on Enterprise Buildings Integrator, a platform used in more than 30,000 buildings.
- A Forge Cognition deployment at NVIDIA's Santa Clara HQ cut unnecessary alerts by an average of 90 per cent and found up to 10 per cent potential energy savings.
- System Sensor Distinguish detector technology distinguishes fire signatures from steam, dust and mist to reduce nuisance alarms.
- Niagara 5 connects HVAC, fire, security, access control, lighting and energy systems across vendors.
Honeywell Technologies has launched a suite of AI-driven building automation and fire safety products, headlined by a command-and-control platform called EBI One and a detector designed to tell the difference between real fires and steam, dust and mist. The company unveiled the releases at its Accelerate 2026 customer event on 12 November 2026.
The announcement spans three of Honeywell's building-focused businesses: building operations, fire and life safety, and security. Together, the products form the most visible snapshot yet of a strategy the company has been building since it introduced Forge Cognition, an agentic AI and simulation platform developed with NVIDIA and Dell Technologies.
The stakes are practical, not abstract. Nuisance fire alarms are a persistent problem for facilities teams and emergency responders, and Honeywell's own deployment data suggests AI can cut the noise dramatically. At NVIDIA's headquarters in Santa Clara, a co-creation deployment of Forge Cognition cut unnecessary alerts by an average of 90 per cent and identified potential savings of up to 10 per cent in energy and utility consumption.
What is EBI One?
At the centre of the launch sits EBI One, a new command-and-control platform for building operators. It runs on the intelligence layer of Honeywell's Forge Cognition platform, which continuously assesses building conditions and guides what operators should do next.
EBI One is not a clean-sheet product. It builds on Honeywell's Enterprise Buildings Integrator platform, which the company says is already used in more than 30,000 buildings. That installed base matters. It means Honeywell can push AI capabilities into a huge fleet of existing commercial buildings rather than selling a rip-and-replace upgrade.
The new platform includes three headline capabilities:
- Conversational AI, letting operators interact with building systems in natural language.
- Alarm assistance, aimed at helping operators triage and respond to alerts.
- Open integration with other building systems, connecting data that traditionally lived in separate silos.
How does Honeywell plan to cut false fire alarms?
The fire and life safety updates focus on earlier detection and faster, better-informed incident response. One addition is low-level carbon monoxide monitoring. It gives operators immediate visibility of CO concentration levels plus trend analysis, so they can watch a problem develop before it becomes an emergency.
The second is the one facilities teams will notice most. Honeywell is introducing System Sensor Distinguish detector technology, engineered to distinguish between actual fire signatures and conditions like steam, dust and mist. The goal is straightforward: fewer nuisance alarms.
Why that matters is easy to overlook. Unwanted alarms drain responder attention, erode trust in alarm systems and pull skilled staff away from real problems. In a building portfolio running tens of thousands of sites, even a modest reduction in false alerts compounds quickly.
Honeywell is also launching Incident Command, a software application that pulls incident information, accountability data, operating procedures and live information from connected buildings into a single view. The intent is to help responders track changing conditions while an incident is still unfolding, rather than piecing together the picture afterwards.
What did Honeywell's CEO say?
Juan Picon, President and CEO of Honeywell Technologies Building Automation, framed the launches around a labour problem. Building owners and operators, he said, are managing more complex environments with fewer skilled staff available.
"Buildings are at an inflection point, with owners and operators being asked to manage increasingly complex environments, often with fewer skilled resources," Picon said.
His second point cuts at the industry's data glut. "Customers today need more than connected systems and volumes of data. They need technologies that turn information into actionable intelligence, simplify operations and enable better decisions."
That argument explains the shape of the announcement. Every product on the list — EBI One, Distinguish, Incident Command, the security suite, Niagara 5 — is about filtering and interpreting signals, not just collecting them.
What's coming to security systems?
Alongside the products available now, Honeywell outlined further work in building security software. The LenelS2 Adaptive Intelligence Suite will add a layer of AI capabilities on top of existing security platforms:
- AI-weighted alarm management
- Anomaly detection
- Natural-language reporting
- Guided workflows
- Unified dashboards
The design choice here is deliberate. The suite sits on top of current installations, starting with LenelS2 OnGuard, so customers can upgrade their operations without replacing existing systems. Honeywell says the functions are designed to reduce alarm fatigue and improve situational awareness for security teams — the same philosophy of cutting noise that drives the fire safety releases.
What is Niagara 5?
Honeywell also detailed the next generation of its Niagara framework, called Niagara 5. The framework connects systems across heating, ventilation, air conditioning, fire, security, access control, lighting and energy — including equipment from different vendors.
That vendor-neutral scope is the framework's core value. Commercial buildings rarely run on one manufacturer's kit, and integration work across mixed systems has traditionally been slow and expensive.
Combined with Forge Cognition, Niagara 5 is intended to reduce engineering and commissioning work for three groups: building owners, operators and system integrators. For integrators in particular, less commissioning time means faster projects and lower delivery costs.
Why this points toward autonomous buildings
Picon tied the whole announcement to a longer arc. Honeywell sees AI and automation as part of a wider shift in how buildings are managed, and it is positioning itself ahead of that shift.
"At Accelerate 2026, we're demonstrating how Honeywell Technologies is combining decades of building automation expertise with AI and advanced technologies to help customers address these challenges and accelerate the transition to autonomous building operations," he said.
The Santa Clara numbers give that ambition its clearest evidence point. A 90 per cent average reduction in unnecessary alerts, plus a potential 10 per cent cut in energy and utility consumption, is the kind of result that turns AI from a demo into a procurement argument.
The open question is deployment speed. Honeywell now has the platform layer (Forge Cognition), the operator interface (EBI One), the detector hardware (Distinguish), the responder tool (Incident Command), the security add-on (LenelS2 Adaptive Intelligence Suite) and the integration backbone (Niagara 5) — a full stack aimed at buildings that increasingly run themselves. How quickly the 30,000-building Enterprise Buildings Integrator base adopts it will be the metric to watch.
Source: itbrief.com.au
More from Elena Vasquez
Show full bio
Market editor covering technology and product news in artificial intelligence and machine learning at AI In Context.
524 articles
Related articles
- Anthropic Warns Small AI Models Could Self-Replicate Through IoT Devices
- Red Hat ships OpenShift 5 with AI agents, five-year lifecycle
- OpenAI's Long-Horizon Model Broke Out of Its Sandbox to Post to GitHub
- Illumio Joins WWT's ARMOR Framework for AI Breach Containment
- New Relic's Autopilot Kubernetes Agent Hits Public Preview