Smart Buildings

IS 2189:2026 Makes Fire-Alarm Health Monitoring an Option

The fifth revision of IS 2189, published by the Bureau of Indian Standards in March 2026, defines “online health monitoring of equipment” as IoT technology that watches the health of fire safety systems from cloud services, with data available online anywhere.[1] It also lists IoT health monitoring as one of three ways to service a system, described as helping facility managers when no site employee is present.[1] That is the article’s central argument: the standard lets a building design fire-alarm uptime in and watch it, but as an option. Nothing we found makes it a requirement, so a tender has to ask for it.

The sections below set out where the 2026 text mentions monitoring, how it words each mention, what it says about wireless detection and the fire command centre, and what a specification still has to settle. Clause numbers refer to the BIS text as hosted by NSC India. We read the clauses ourselves; we have not compared them line by line with the 2008 edition.

Five places where the text points beyond the panel, and how firmly

ClauseWhat the 2026 text saysWording strengthWhat a specification still has to add (ours)
3.56 DefinitionOnline health monitoring is IoT technology that monitors the health of fire safety systems and fire protection equipment from cloud services (web access or mobile application), giving online access to the data anywhere “ensuring corrective action when required”.[1]Definition onlyWhich devices and which data points count as “health” at this site
7.1.2 IoT and M2MTechnologies “may also be adopted”. Connecting the panel to a cloud-based monitoring server to track fire incidents and mean recovery time, and fault occurrences and mean restoration time, “offers a significant advantage over manual monitoring conducted via periodic inspections”.[1]“May”Whether those two time measures are reported, to whom, and how often
7.3.1 ServicingServicing is by a contract with a competent contractor, by a trained employee of the user, or by “IoT based health monitoring of the system to assist facility managers in absence of site employee”, which “shall provide the real time health of the system”.[1]One of three routesWho receives alerts if option (c) is chosen, and who must act on them
7.4.5 (g)Where online monitoring exists, quarterly-test records can be kept on the cloud for access to past records “without manuals”.[1]PermissiveRetention period, export format and who may retrieve records
4.7 Fire command centreA fire command centre is shown with mimic panels and graphic control software, “or alternatively, IoT based remote monitoring system may be provided”, which may be integrated with the local fire brigade if available.[1]“May”Whether remote monitoring supplements or replaces the staffed room
The foreword lists an informative Annex D on online health monitoring among this revision’s major changes.

That Annex D is informative, so it guides rather than binds. It describes a system that gathers the working condition of the fire alarm (fault or alarm state), of hydrant systems (main-line pressure, pump on and off records, trip status) and of extinguishers (pressure, location, obstruction), and uploads the data to a monitoring centre.[1] It also says lifespan records let people act when events repeat, and that notifications can be set up in a hierarchy with escalations.[1] The annex does not say how long such data should be kept or who should see it.

We searched the text for any wording on encryption, authentication, passwords or data retention for the cloud link and found none. A building that connects its life-safety panel to a cloud service therefore has to set those terms itself.

The code already defines uptime through checks that a log book records

Monitoring matters because the standard already asks for a rhythm of checks. A daily check confirms that the panel indicates normal operation, that any fault is entered in the log book and attended to urgently, that faults logged the previous day received attention, and that the panel has not been left non-functional by a burnt fuse, a faulty power supply or inadvertent switching off.[1] A weekly check confirms that the week’s events were addressed correctly.[1] The quarterly check covers log-book entries, batteries, the alarm function in each zone, any link to a remote manned centre, and a visual check that at least a 750 mm radius below every detector is clear.[1]

Two other provisions turn on time. For addressable systems, all trouble signals must reach the control panel within 10 seconds of occurring.[1] And if an inspection leaves a deficiency uncorrected, the system owner or the owner’s representative must be told of the impairment in writing within 24 hours.[1] For hospitals, hotels and old people’s homes, an engineer must be on call at all times, with a telephone request for emergency service executed within 24 hours.[1]

Our reading is that these are exactly the events a monitoring service could time-stamp. A paper log can show that someone looked at the panel. A time-stamped record can show how long a fault stayed open. The standard stops short of requiring that second record, which is why the choice sits with the owner.

Conceptual diagram of a fire-alarm health data path from detectors and panel through an IoT link and cloud service to alert recipients and records, with a specification question under each step. Not a product design.
Conceptual only: this is our illustration of the questions at each step, not a design from IS 2189:2026 or any vendor.

Wireless detection is allowed, but the text asks for proof of supervision

Clause 4.9 covers wireless fire detection and alarm systems that use radio communication for fire, fault and supervisory signals. Systems may be fully wireless or hybrid. A hybrid system must give supervision, fault monitoring and alarm integrity “equivalent to fully wired systems”.[1] The text calls for reliable and secure communication over radio paths in normal and adverse conditions, plus fault tolerance, redundancy and supervision so that an alarm still gets through if a link fails, suffers interference or a component malfunctions.[1]

Verification is specific. It includes communication integrity supervision, resistance to temperature, humidity and vibration, power supply performance including battery life, and immunity to electromagnetic interference.[1] Conformance assessment is required, and test documentation and certification must be made available to the authority having jurisdiction on request.[1] The standard lists licence-exempt bands such as 865 to 867 MHz, 868 MHz and 2.4 GHz, and puts the burden of WPC approvals on the manufacturer or OEM.[1] We quote the band list from the standard and have not checked it against the current WPC notification.

The listed applications are heritage and protected buildings where cabling is restricted, temporary structures, retrofit and renovation projects, and large campuses where cabling is impractical or uneconomic.[1] For anyone retrofitting a building, the practical point is that the certification file, not the radio brand, is what an inspector may ask for.

A fire command centre is a staffed room, so an alert needs a person at the end of it

Clause 3.21 defines a fire command centre as a room permanently staffed 24 hours a day, seven days a week, by qualified and trained fire personnel, within or near the premises. It is equipped to show the situation in each protected premises and has the communications needed to call for help, including two-way communication with fire exit staircases and refuge areas.[1] Clause 4.7 adds that the control room should display the latest fire alarm drawing, visible to a trained fire officer.[1]

Remote monitoring changes the design question. If the alert goes to a phone, someone has to be named to answer it at 3 a.m. If it goes to a staffed room, the room needs a defined feed. The standard says remote monitoring may be integrated with the local fire brigade, if available. It does not say who must acknowledge an alert, and we do not read it as making any building change its staffing.

Ten items to settle before a tender says “IoT health monitoring”

This checklist is ours, not policy and not part of IS 2189:2026. It turns the clauses above into questions a bidder must answer in writing.

  1. Scope of health. Name the panels, loops, detectors, modules, batteries and sounder circuits the service must report on, and the states it must capture: alarm, fault, supervisory, disablement.
  2. Read-only link. State whether the monitoring link can only read from the panel. If it can send commands, that is a different cyber risk and needs separate approval.
  3. Cyber review. The standard is silent on authentication and encryption for the cloud link, so require them in the contract and review the connection as you would any building system. Our notes on BMS design for a control room and on CERT-In’s AI agent alert and smart buildings cover the wider building-network risks.
  4. Alert routing. List who is alerted for each state, how, and who must acknowledge within what time.
  5. Recovery measures. If you want the incident and fault recovery times that clause 7.1.2 mentions, say how they are calculated and reported.
  6. Records. Fix retention, export format and who may retrieve records for an inspection or audit.
  7. Account and data ownership. The owner, not the installer, should hold the administrator account and be able to take the data with them.
  8. Outage behaviour. Ask what happens to alerts and records if the internet link or the cloud service fails, and whether the panel works normally without it.
  9. Wireless file. If any wireless devices are used, ask for the conformance and test documentation, RF supervision details and WPC approvals before award.
  10. Human backup. Keep the daily check, the log book and the named engineer. The standard treats monitoring as an aid, and an alert nobody owns protects no one.

Trisim’s view: put the monitoring decision in the design brief, not the maintenance renewal

Trisim’s view is that the standard gives owners the vocabulary to ask for fire-alarm health data, and the cheapest moment to ask is before the panel is specified. Added later, it means a second integration, a second contract and a second cyber review.

Related reading: designing a fire alarm system as per codes and how to design a BMS for a control room. For enforcement, see our piece on Odisha’s fire-inspection results.

Disclosure: this article draws only on public documents. Nothing came from BIS, any manufacturer or any building owner, and none has reviewed it. The banner below is a Trisim Global Solutions promotion for Honeywell gas suppression solutions. It is unrelated to the standard discussed here, and we make no claim of any Honeywell partnership.

Protecting a data centre or critical equipment room? Ask Trisim Global Solutions about Honeywell gas suppression solutions.

Honeywell gas suppression solutions for data centres - enquire with Trisim Global Solutions

Sources

  1. Bureau of Indian Standards, IS 2189:2026, Selection, Installation and Maintenance of Automatic Fire Detection and Alarm System, Code of Practice (Fifth Revision), copy hosted by NSC India, March 2026
  2. Ahuva (fire-safety vendor), NBC 2016 and NBCS 2026: what India's building code says about fire detection and alarm, 2026

Leave a Reply