In a New York apartment building, the front door is a statute.

Not automatically. The state requirement for two-way voice and remote release from the apartment attaches to Class A multiple dwellings erected or converted after 1 January 1968 with eight or more apartments. A 1931 walk-up is outside that trigger as built.

The retrofit provision is what usually catches pre-1968 buildings. Where a majority of tenants consent in writing, the owner has to install the system, beginning within 90 days and finishing within six months. So the practical answer is that a pre-1968 building is not required to have one until its tenants collectively ask for it in writing, at which point the clock is short and the scope had better already be understood.

Separately, the self-closing door requirements for residential buildings with three or more units apply regardless of construction date, and that is a different obligation with a different enforcement program behind it.

Whatever the inspector recorded as defective, physically working again. The trap is that owners fix the visible half. If the write-up concerns the entrance door, a new closer does not help if the latch still fails to engage the strike, because the door is then self-closing and not self-locking. If it concerns the intercom, restoring the buzzer does not help if the voice path is dead, because the requirement is two-way communication and remote release together.

Practically, these get written as hazardous conditions with a short correction window and the clock does not care whether you have a contractor lined up. That is the reason to survey the entrance before you have a violation rather than after: the diagnostic visit is the slow part, and a building that already knows its riser condition can be scheduled inside a correction window instead of racing it.

We document what we found, what we changed and what it tests at, with photographs, so you have something to submit rather than a paid invoice and a hope.

It depends entirely on where the equipment lives. A lobby panel, a door station, an electric strike, a power supply and a wireless or app-based release path can all be done without entering a single unit. Replacing in-unit handsets or adding in-unit video monitors obviously cannot.

This is one of the strongest practical arguments for a system where the resident’s answering device is their own phone: the in-unit half of the job disappears, along with the notices, the repeat visits and the four apartments that never answer. The trade-off is honest and worth stating: a phone-based system depends on the building’s network or on cellular service and on the resident having and charging a phone, so the design needs a documented fallback for outages and for residents who will not use an app. In most buildings that means keeping a hardwired release at the super’s station and a physical key path that does not depend on anything electronic.

On an old 125 kHz proximity system, more than most owners think, not because that one fob is dangerous, but because it can be copied trivially and because on many of those systems there is no practical way to revoke a single credential. Plenty of legacy installations were set up with one shared facility code and no per-credential record at all, which means the only way to remove one fob is to re-badge the building.

On a properly administered modern system it is a thirty-second job in a browser, done by your own staff, with a log entry showing who did it and when. That difference, not the reader on the wall, is the actual product you are buying.

If you are on a legacy system and re-badging is not in this year’s budget, the interim measure that helps most is a door position switch and a forced-door alarm on the entrance, so at least the building knows when a door is opened without a valid credential or held open.

We are not the right people to answer that, and we would rather say so than guess. Cost recovery in rent-regulated units follows the housing agency’s own procedures and paperwork, and it depends on the regulatory status of the units, the nature of the improvement and the current rules, which change. That is a question for your managing agent and your counsel, before the work rather than after.

What we can do is give you the input they need: an itemized scope separating building-wide equipment from per-unit equipment, separating replacement from betterment, and separating the parts that are legally required from the parts that are elective. A single lump-sum number is very hard for anyone to work with in that conversation.

Sometimes it is the smartest thing in the building, and sometimes it is the reason the new system never quite works. Keeping a legacy release path at the super’s station is genuinely useful: it means a network outage or a platform problem does not leave anyone stranded on the sidewalk. Keeping a whole parallel intercom energized on the same riser is different: two systems sharing conductors is a reliable way to generate hum, crosstalk and intermittent faults that take a day to chase.

The version we normally recommend is a hardwired secondary release, a documented mechanical override, and the old head end de-energized and left in place rather than removed. That preserves the fallback, removes the interference, and keeps the option of reverting if something about the new platform turns out not to suit the building.