Smart Intercom, Access Control and Elevator Emergency Communication: The Phone in the Lift Forced the Convergence

Three products, three suppliers, three budgets: the door intercom, the credential reader that opens the door, and the two-way voice device inside the elevator car. Only one of them is a life-safety system. All three are now running on the same network — not because anyone set out to converge a building’s entry layer, but because the copper line underneath the least expensive of the three is being retired from under it.

The cheapest device, and the only one that must work

A door intercom fails and somebody buzzes twice. A credential reader fails and somebody reaches for a key. The two-way communication device inside an elevator car fails and a trapped passenger has no way to summon help at all. That asymmetry — small price, absolute requirement — explains most of how this market behaves, and why the least glamorous of the three products ends up setting the timetable for the other two.

It also explains why they are so rarely bought together. The lobby intercom belongs to the developer’s low-voltage package or the security integrator. The credential reader belongs to the facilities or security team. The car phone belongs to the elevator contractor, because it is treated as part of the elevator and is serviced by elevator personnel. Three line items, three suppliers, three refresh cycles — and, until recently, three different kinds of wire.

Convergence is an infrastructure fact, not a strategy

Access control spent two decades migrating from standalone controllers to networked systems and then to the cloud, so that the credential decision increasingly happens in software. Intercom moved from four-wire analogue audio to IP devices running video over Power over Ethernet, so that the door station is now a network endpoint with a camera in it. The elevator phone moved from a dedicated copper pair dialling a central station to an IP or cellular device doing the same job over a path it does not own.

The three converge on the same switch, the same PoE budget and often the same VLAN. That is the substantive change: the entry layer of a building has become an IT system with a life-safety device inside it. The specification question stops being which intercom and becomes what happens to the door, and to the lift, when the network does not work. Those are not the same question, and they are answered by different people.

The lift phone is not a phone. It is a code artefact

The American Society of Mechanical Engineers published the twenty-fourth edition of its elevator and escalator safety code, ASME A17.1-2025/CSA B44:25, with a date of issuance of 30 January 2026. Section 2.27, Emergency Operation and Signaling Devices, is the part that governs the device in the car, and what it requires is much more specific than “a phone”.

Permanent two-way voice communication must exist between the elevator car and a location staffed by authorised personnel who can take appropriate action. If the call is not acknowledged within 45 seconds, it must be automatically directed to an alternate on- or off-site location. A push button marked with the telephone symbol initiates the call, and once established the connection may be disconnected only when authorised personnel terminate it or a timed termination occurs — and a timed termination requires a voice notification to those personnel at least three minutes in, who then have 20 seconds to extend the call. Where the elevator rise is 18 m (60 ft) or more, the building must also provide a permanent two-way voice communication means at the designated level for emergency personnel, one that can reach each car individually, requires no intervention from inside the car, and overrides calls to outside the building. And if the device normally draws on the building power supply, it must automatically transfer to standby or emergency power.

None of that is a product feature. It is a specification with a clock, an answering location and a power path attached — and two of those three components are not the device. The 2025 edition also tightened the emergency communication requirements so that they work for passengers who are hearing impaired, which is the code admitting that a voice-only path was never the whole requirement.

The jurisdiction is split, and nobody files the paperwork

NFPA 72, the National Fire Alarm and Signaling Code, does govern a family of related systems — rescue-assistance two-way emergency communications for stairwells, areas of refuge and elevator landing or lobby areas — and requires those to be listed to UL 2525 with defined levels of pathway survivability. The communications inside the car are not NFPA 72’s business. The code removed elevator emergency communications from its body in the 2019 edition and left them in the annex, on the reasoning that ASME A17.1 defines and specifies that system and it is serviced by elevator personnel.

So one ground-floor incident touches the fire alarm contractor, the elevator contractor, the security integrator, the voice-and-data installer and the IT team. Every one of them can install the box. None of them naturally owns the record that the box still answers. The scarce asset in this market is not the handset — it is the evidence that the handset still works, which is a filing problem dressed up as a hardware problem.

The copper is leaving the building

The forcing function is not demand. It is the retirement of the analogue infrastructure the device was designed around. In the United Kingdom, the telecommunications industry is switching off the Public Switched Telephone Network by January 2027, and government guidance is explicit that devices reliant on it — naming door entry systems among them — require upgrading to digital alternatives. Ofcom, the sector regulator, reported a 45% increase in the number of PSTN incidents in 2024, with more than 2,600 major incidents on the network. This is a network failing while it is still in service.

In the United States, AT&T said in 2024 that it plans to exit its copper network across the majority of its wireline footprint by the end of 2029, and began discontinuing legacy copper-based services in parts of that footprint in the summer of 2026, phased over several years, with the commitment that no customer loses access to home phone or 911. Reporting at the time confirmed the 2029 target and noted that the carrier developed a POTS replacement capable of running over fibre or wireless. The same dynamic applies at every major carrier.

A device that sat behind a wall for thirty years, wired to a line that always worked, is now an item on an IT migration plan with a life-safety classification. And its failure mode is silent. The phone in the car looks perfectly fine until somebody presses the button, which is why the impaired-system rule matters: a fire alarm panel that loses its communication path is classified under NFPA 72 as an impaired system, and in most jurisdictions that triggers notification to the authority having jurisdiction and a fire watch until the system is restored. The compliance exposure arrives before anyone notices the fault.

Accessibility law has moved to the answering end

The European Accessibility Act applied from 28 June 2025, and its reach is instructive. Ireland’s Commission for Communications Regulation is designated as a compliance authority for electronic communications services and, separately, for the answering of emergency communications to the single European emergency number 112. Providers must offer Real Time Text and Total Conversation, including in emergency calls.

That is a requirement about the path on the other end of the call, not about the button on the wall. It shifts the unit of compliance from the handset to the service — and it is the direction of travel in the code as well, where the 45-second non-acknowledgement redirect already assumes a second answering location exists and is staffed.

Follow the money: the fleet, not the box

The elevator side of this market is not a new-equipment business, and the numbers say so plainly. Otis reported that the global installed base reached 23 million units at the end of 2025 and continues to expand at a mid-single-digit pace, that it maintains approximately 2.5 million units — the largest portfolio in the industry, about 10.9% of that base (Pheonix Research calculation: 2.5 million ÷ 23 million) — and that those units carry roughly 2.5 billion passengers a year.

Its 2025 results show where growth sits. Net sales were $14.4 billion with flat organic growth, but service net sales were $9.4 billion, up 6% and up 5% organically — roughly two-thirds of group sales (Pheonix Research calculation: 9,442 ÷ 14,400 = 65.6%). Maintenance and repair grew 4% organically and modernisation grew 9%; modernisation orders rose 43% in the fourth quarter, and the modernisation backlog ended the year up 30% against a new-equipment backlog up 2%, with new-equipment organic sales down 7%.

Schindler’s 2025 revenue was CHF 10,947 million, up 1.3% in local currencies, with operating profit of CHF 1,384 million at a 12.6% margin. KONE reported 2025 sales of EUR 11,245.2 million, up 1.3% and up 4.0% at comparable exchange rates, with adjusted operating income of EUR 1,369.3 million, a 12.2% margin. Three different companies, one shape: flat new equipment, growing installed base, and the margin in service.

Modernisation is also the demand event for the device in question, because it is the moment the car is open and the wiring is exposed. That replacement cycle is now being pulled forward by a network deadline rather than a buying decision — which is a very different thing to plan for.

The security layer is growing faster than the buildings it sits in

The access and entry layer above it is growing faster than construction. ASSA ABLOY reported 2025 sales of SEK 152,409 million, up 1% in total and up 3% organically, with fourth-quarter organic growth of 4% and a record fourth-quarter operating margin of 16.8%. Its Global Technologies division, which houses HID, grew 9% organically; in the quarter, electromechanical sales grew 8% organically within the regional divisions, while the Americas grew 5%, EMEIA 4% and Entrance Systems 2%, and Asia Pacific fell 2%. In October 2025 the group acquired Kentix, a German maker of monitoring and access control products for data centres.

Allegion’s fourth-quarter 2025 results show the same split from the other end: Americas revenues up 6.1%, or 4.8% organically, with the non-residential business up high single digits organically and the residential business down high single digits. dormakaba, whose financial year ended 30 June 2025, reported net sales of CHF 2,870.1 million, organic growth of 4.1% driven by volume of 2.4% and price of 1.7%, an adjusted EBITDA margin up 80 basis points to 15.5%, and net profit of CHF 188.0 million — and agreed to acquire TANlock, a German access solutions company focused on data centres, at the start of its new financial year.

Note the pattern rather than the individual deals. Two of Europe’s three largest access-control groups bought data-centre access businesses within twelve months of each other. The vertical where a door credential carries the most value is not the office lobby or the apartment entrance. It is the hall outside a server room — which is also precisely where the elevator-code logic, the network-resilience question and the access-control business all collide.

Two supply bases, one specification page

There is a geopolitical seam running through this market as well. The FCC added video surveillance and telecommunications equipment produced by Hangzhou Hikvision, Dahua Technology and Hytera Communications to its Covered List on 12 March 2021, covering equipment used for public safety, the security of government facilities, physical security surveillance of critical infrastructure and other national security purposes. Since 6 February 2023, equipment identified as covered has been barred from obtaining an FCC equipment authorisation, and the Commission has run a separate process to stop further importation or marketing of already-authorised covered equipment — with a June 2026 public notice covering equipment added to the list in 2024 or earlier.

Hikvision’s own results show how much of the entry layer that supply base touches. For 2025 the company reported revenue of RMB 92.51 billion (USD 12.95 billion), up 0.01%, and net profit of RMB 14.20 billion (USD 1.99 billion), up 18.52%. It singled out its non-video categories — access control, alarm systems and commercial displays — as a growth engine, and its “innovative business” contributed RMB 25.45 billion, or 27.51% of revenue.

So the price-setting layer of the global entry market is a Chinese vendor group that is partly fenced out of federally funded US buildings, while the European listed layer grows mid single digits with its margin in software and data-centre doors. Both are selling into the same specification pages, and increasingly the same buildings.

Where Pheonix Research covers this

Pheonix Research does not publish a market called “smart intercom”, “access control” or “elevator emergency communication”, and any single number presented as one should be interrogated for what sits inside it — devices only, or installation; hardware only, or monitoring; the intercom alone, or the whole entry layer. What Pheonix Research does cover is the built environment these systems are bolted into and the safety systems they sit beside. The [Construction & Infrastructure industry](https://www.pheonixresearch.com/construction-infrastructure/) and its [Infrastructure segment](https://www.pheonixresearch.com/construction-infrastructure/infrastructure/) are where the installed base and the retrofit cycle live. The closest adjacent budget line is fire protection: Pheonix Research sizes the [Global Fire Protection System Market](https://www.pheonixresearch.com/construction-infrastructure/infrastructure/market-report/global-fire-protection-system-market/) at USD 78.65 billion in 2025, reaching USD 129.84 billion by 2033 at a 6.46% CAGR — a market that shares the same code path, the same impaired-system rule and, increasingly, the same network as the entry layer described here.

What to watch

Three things will decide how quickly this convergence actually happens, as opposed to how often it is announced.

The first is the deadline stack. The UK PSTN goes off in January 2027 and AT&T’s copper is largely gone by the end of 2029, and both dates are immovable in a way that equipment refresh cycles are not. The constraint on meeting them will be qualified integrators and staffed answering capacity, not handsets — because the answering location is the part the code actually specifies, and it is the part nobody sells as a box.

The second is how far up the chain the accessibility requirement travels. It has already reached the emergency answering service in Europe and the alternate answering location in the elevator code. The unresolved question is liability: when the network carrying the life-safety call is the same network carrying the cameras, the door and the guest Wi-Fi, the organisation that owns the consequence is rarely the organisation that owns the switch.

The third is whether the elevator code absorbs what the data-centre door has taught the security industry — continuous verification, network-native credentials, remote diagnostics — the way the 2025 edition absorbed the hearing-impaired requirement. The installed base is 23 million units and growing at a mid-single-digit pace. Every one of them has a device in it that must answer.

That is the whole point. A building does not converge its entry layer because convenience demands it. It converges because a line was retired underneath the only part of it that was ever allowed to matter.

Frequently Asked Questions

What is smart intercom, access control and elevator emergency communication?

It is a building's entry-and-egress layer: an IP-based intercom at the door, a credential reader that decides whether the door opens, and the two-way voice communication device required inside every elevator car. The first two are security and convenience products. The third is a life-safety device governed by the elevator code, and it is the one that sets the compliance timetable for the other two.

Are elevator emergency phones required by code?

Yes. ASME A17.1/CSA B44, Section 2.27 on emergency operation and signaling devices, requires permanent two-way voice communication between the elevator car and a location staffed by authorised personnel, with an automatic redirect to an alternate on- or off-site location if the call is not acknowledged within 45 seconds, automatic transfer to standby or emergency power where the device normally runs on building power, and — where the rise is 18 m (60 ft) or more — a separate two-way voice communication means at the designated level for emergency personnel. The edition issued on 30 January 2026 (ASME A17.1-2025/CSA B44:25) also tightened the emergency communication requirements so that they work for passengers who are hearing impaired.

Why are buildings replacing intercom and access control now?

Two forces. First, the legacy analogue network underneath the entry layer is being retired — the UK PSTN is scheduled to be switched off by January 2027, and AT&T has said it plans to exit copper across most of its wireline footprint by the end of 2029. Second, the security layer is growing faster than the buildings it sits in: access control vendors are reporting faster growth in software, electromechanical hardware and data-centre access than in the wider building market, and elevator growth is concentrated in service and modernisation rather than new equipment.

Who supplies this market?

Three distinct supply bases. European listed access and entry groups — ASSA ABLOY, Allegion and dormakaba — supply credentials, door hardware and entrance systems. Specialist intercom vendors such as Aiphone, Comelit, Fermax and 2N (Axis) supply the door station. Elevator original equipment manufacturers including Otis, Schindler and KONE supply the car communication device as part of their maintenance and modernisation business. Chinese vendors including Hikvision and Dahua are major in video surveillance and intercom hardware but face US restrictions on equipment authorisation under the FCC Covered List.