Short answer: You can add facial recognition to existing access control in four ways: integrate Edge.AI by Proptech.AI with your current panels via Wiegand, run it in parallel, mix the two in a hybrid setup, or retire the legacy system and run Edge.AI standalone. Because the doors and turnstiles stay, and in most modes the panels too, a retrofit is the most cost-effective proptech path to hands-free entry.
Key takeaways
- Integrated mode links Edge.AI to existing panels over Wiegand, so the current system stays in place and gains facial recognition, QR codes and analytics.
- Parallel mode lets you prove adoption at one entrance while the legacy system carries on untouched.
- Hybrid mode fits mixed properties, such as integrated lobby turnstiles plus standalone amenity doors.
- Full replacement suits a legacy platform near end of life, because Edge.AI is a standalone access control system.
- Reusing doors, turnstiles and panels moves budget from replacing working hardware to adding capability.
Why retrofit instead of ripping out a working system?
Most of what a rip-and-replace would tear out still works, and replacing turnstiles, doors and panels pays again for capital the building already spent. A retrofit spends money only where something changes: the moment a person proves who they are.
That is why, among proptech investments, adding AI to an existing access system delivers so much capability per dollar, and why Edge.AI facial recognition access control offers four deployment modes rather than one.
What are the four ways to add facial recognition to an existing system?
The four modes are integrated, parallel, hybrid and full replacement, and each keeps a different share of the legacy system in place.
1. Integrated via Wiegand
Here Edge.AI talks to the access control panels you already own over Wiegand, the long-established interface between readers and panels. Your current system stays in charge of cardholders, schedules and reports, Edge.AI contributes facial recognition, mobile QR codes and analytics, and existing cards keep working on the readers you keep.
2. Parallel
Edge.AI runs alongside the existing system, for example during a pilot or a phased rollout, so you can judge enrollment, entry flow and analytics on real traffic while the legacy system carries on.
3. Hybrid
Hybrid mixes both approaches across one property: Edge.AI integrated at lobby turnstiles, where the card system already lives, and standalone on amenity doors such as a fitness center or roof terrace.
4. Full replacement
Edge.AI is a standalone access control system, so it can replace the legacy system entirely. Doors and turnstiles stay; the legacy software is retired.
Which mode fits your building?
Choose by the condition of what you have and how much change occupants can absorb at once.
| Mode | What stays in place | What Edge.AI adds | Best fit when |
|---|---|---|---|
| Integrated via Wiegand | Panels, software, cards and schedules | Facial recognition, QR codes and analytics | The current system works and is staying |
| Parallel | The entire legacy system | A separate trial at chosen entrances | You want proof on real traffic before committing |
| Hybrid | Legacy control at some entrances | Integrated and standalone zones | Entrances differ in age, hardware or use |
| Full replacement | Doors and turnstiles | The whole access, visitor and analytics platform | The legacy system is near end of life or fragmented |
Where Edge.AI is hosted
Both cloud and on-prem hosting are available. The cloud suits portfolios that want one centralized view across sites, while on-prem suits owners and tenants whose policies require access data to stay on infrastructure they control.
Will Edge.AI work with the panels you already have?
If your panels accept Wiegand input, the long-standing norm for card readers, integrated mode is available. These approaches work with systems built on common controller platforms, and our access control system comparisons explain how Edge.AI sits next to each. For LenelS2 software or Mercury Security hardware, see Proptech.AI vs. LenelS2 and Proptech.AI vs. Mercury Security, and confirm compatibility for your exact panels with Proptech.AI.
Retrofits also raise a security question at the wall. Rip an all-in-one biometric reader off its mount and you may reach the building network, the lock relays or the cable to the panel, and the device leaves with the user data on it. Edge.AI splits the hardware so those connections and data sit away from the unit people face.
What does a retrofit migration look like, step by step?
A retrofit migration moves one entrance at a time, with the policy work done before anyone enrolls.
- Inventory doors, turnstiles, elevators, panels and credential types.
- Assign a mode to each entrance and choose hosting.
- Write the biometric policy, consent notice and opt-out path (QR code or RFID card).
- Pilot at one entrance, enrolling a first group through emailed selfie links.
- Review entry flow and alerts, then extend to other entrances, retiring legacy parts at end of life if full replacement is the goal.
How much does reusing existing hardware save?
The savings come from what you do not buy.
Illustrative example: Assume an office building with 6 turnstile lanes, 40 card-reader doors on a working panel system and 1,200 badge holders. Every figure is an assumption, so substitute your own quotes.
- Assume replacing the turnstiles would cost $15,000 per lane: 6 × $15,000 = $90,000.
- Assume new controllers and readers at $2,500 per door, installed: 40 × $2,500 = $100,000.
- Assume re-badging everyone onto a new card system at $8 per card plus 10 minutes of desk time at $30 per hour: 1,200 × $8 = $9,600, and 1,200 × 10 ÷ 60 × $30 = $6,000.
Under these assumptions, a rip-and-replace puts $90,000 + $100,000 + $9,600 + $6,000 = $205,600 of hardware and re-badging on the table. An integrated or hybrid retrofit keeps those turnstiles, doors and panels in service. Weigh that avoided spend against a subscription quote from Proptech.AI for the entrances you actually plan to change.
How did 101 Greenwich and 330 Madison add facial recognition?
In both, Edge.AI was layered onto a card-based system the building already ran. 330 Madison in New York City is a JLL-portfolio tower of 39 floors and 855k RSF, and hundreds of people move through it each minute. Proptech.AI tied facial recognition and appless self-enrollment into its 6 turnstiles and its existing card-based access control, keeping access secure 24/7.
101 Greenwich, a JLL-managed property with 1,500 tenants across 25 floors and 480k RSF, previously depended on headcounts done by hand and a limited access control setup. One Edge.AI sensor went onto each of its six turnstiles and was tied into the existing card-based system, giving the building 24/7 secure access. More for office owners: commercial real estate access control.
What should you check on privacy and accuracy before enrolling anyone?
Check consent, alternatives, data separation and the evidence behind any accuracy claim. In Edge.AI, the biometric privacy consent opens enrollment, anyone who opts out can use a QR code or RFID card, any person’s data can be deleted on request, and PII lives apart from both the device and the management console.
Proptech.AI specifies 99.9% accuracy for its multi-sensor device, backed by fraud and spoofing alerts. Whichever vendor you evaluate, ask how its accuracy figures were established: the FTC, in its May 2023 policy statement on biometric information, cautions that accuracy claims without substantiation can be deceptive, and NIST’s FRTE 1:1 evaluations offer a reference point for independent algorithm testing. This is general information, not legal advice.
Frequently asked questions
Yes, facial recognition can be added to an existing access control system without replacing it. Edge.AI by Proptech.AI can link to the panels a building already uses through Wiegand, leaving the current system running while it adds facial recognition, mobile QR codes and analytics. A building can also trial Edge.AI in parallel, combine integrated and standalone doors, or move entirely to Edge.AI as its only access system.
Wiegand is the long-established interface that connects card readers to access control panels, and many existing panels accept it. Because Edge.AI by Proptech.AI can connect to existing panels via Wiegand, a building can add facial recognition at its current doors and turnstiles while keeping its panels, cardholder records and schedules. Owners should confirm compatibility for their specific panels with Proptech.AI before planning a rollout.
A building should consider full replacement when its legacy access control system is near end of life, when a portfolio runs several incompatible systems, or when the owner wants access control, visitor management and analytics in one platform. Edge.AI by Proptech.AI is a standalone access control system, so it can replace the legacy system entirely while the existing doors and turnstiles stay in place.
Edge.AI facial recognition access control from Proptech.AI runs either in the cloud or on-prem, and the right choice depends on the owner’s IT and data policies. Cloud hosting suits portfolios that want one centralized view across many sites. On-prem hosting suits organizations whose security policies require access data to stay on infrastructure they control, such as some corporate tenants.
Adding facial recognition does not mean everyone has to give up their access card. In an integrated Edge.AI deployment, the existing card system keeps running, so cardholders can keep using their cards while others enroll their face through an emailed selfie link. Edge.AI by Proptech.AI also supports RFID cards and mobile QR codes, and biometric consent is the first step of enrollment, so facial recognition remains a choice.
Map each entrance to the right deployment mode
Share which panels, turnstiles and doors you have, and Proptech.AI can recommend a mode for each. Send the list to sales@proptech.ai or through the contact form at the end of this page with a request for a subscription quote or a demo.
