The first room that betrays a HomeWorks Illumination system is usually not the equipment room. It is the room that just got prettier. New LED trims go in. Millwork gets sharper. Someone expects the old scene button to behave the way it did fifteen years ago, and instead the low end falls apart, the dimming curve chatters, or the keypad still works but suddenly looks older than the architecture around it.

In a Greenwich estate, that moment usually shows up halfway through a larger renovation. Paint is underway. The electrician is swapping loads one room at a time. The owner wants Palladiom keypads, quieter shades, a current app, and maybe a cleaner control layer through Crestron Home. That is when an Illumination job stops being a service call and becomes a migration plan.

HomeWorks Illumination was the predecessor to HomeWorks QS and QSX. It lived in a different control era: separate link types, legacy addressing, RS-232 integrations, and the old Lutron Home Control+ app. Lutron's current support guidance makes that gap clear, and the real upgrade question is no longer whether the old processor can still turn lights on. It is what can still be kept without trapping a renovated house inside legacy behavior. [1]

Why Illumination jobs usually start with one bad room

LED is what exposes the old system

Illumination came from a time when halogen loads were common and the house's control logic was spread across dedicated links and interface boards. Lutron's current App Note 840 still has to explain the old H48 and D48 board world because many of these systems are still in service. [2] The problem is not nostalgia. The problem is modern LED.

Lutron's May 27, 2026 installer guidance reads like a field report from exactly these projects: one room turns out to be LED downlights, another is ELV tape light, another is MLV track lighting, and a decorative fixture still wants incandescent or halogen behavior. That is why the new LED+ Pro Max line is rated up to 250W LED, 500W ELV, 400VA MLV, and 500W INC/HAL. [5] When a single renovated room starts mixing those load types against legacy dimming hardware, the system's age becomes visible fast.

This is also why small stopgap fixes often age badly. Replacing only the failed part can restore light for the moment, but it does not answer whether the rest of the floor is about to follow with the same LED compatibility problems six months later.

The app problem is often what finally forces the decision

Owners rarely ask for a HomeWorks Illumination upgrade because they miss better link architecture. They ask because the house no longer behaves like a current house. Lutron says Illumination can still use the legacy Home Control+ app, but that app is in maintenance mode and is no longer available on Android. [1] That is usually the quietest reason the project stops being optional.

Modern HomeWorks is built around a current mobile and service experience, current shade families, and current processor communication. If the owner expects reliable scene recall, remote support, and a lighting system that does not feel fenced off from the rest of the property, Illumination eventually runs out of credible workarounds.

What usually stays, and what usually does not

Panels can sometimes survive. The processor does not.

A careful survey can save real money here. Lutron's current legacy guidance says QSX condenses the old nine-link Illumination structure down to two primary device families: wired QS and wireless RF. If the house still has legacy RPM panels in workable condition, the Panel Link to QS Link Translator can keep those panels talking to a QSX processor in a near 1-to-1 replacement of the old interface. [2] That is the preserve-first path.

It is not the same thing as preserving everything. Lutron also warns against reusing Illumination 15V power supplies, since QSX devices are powered by 24VDC. It specifically flags older white-label RPMs, generally pre-2003 production, as a risk for strange behavior and outright LED incompatibility. [2] In other words, the right answer is not to keep panels because they are already there. The right answer is to keep only the hardware that still makes engineering sense after the load schedule, fixture schedule, and service plan are honest.

If legacy Grafik Eye hardware is still part of the story, that part is more straightforward. App Note 840 points to current-day replacement with QSGR or QSGRJ models. [2]

Keypad wiring can survive even when the keypads cannot

Illumination keypads look familiar because some of them are the ancestors of devices still on walls today, but they were addressed with DIP switches rather than serial numbers. [1] In App Note 840, Lutron says the legacy keypad wiring can stay and be merged into a QS link, while the keypads themselves move to current families. It specifically calls out Palladiom and SeeTouch as ways to avoid backbox changes. [2]

That is why the aesthetic conversation belongs early. If the owner wants Palladiom keypads in hand-finished metal, this is the time to decide. If a secondary hallway wants a more familiar, less prominent control, current HQRD or HQRA devices may be the better answer. If a multi-location circuit is being rebuilt around Sunnata, Lutron's own guidance says the load side matters: use the HRST-PRO or HRST-8ANS where the circuit requires it. [2] These are not finish-level decisions. They affect how much wall repair, rewiring, and commissioning the project really needs.

Shades are where many budgets get surprised

Lutron's legacy overview is clear: Illumination-era Sivoia QED shades and the Q96 interface do not simply roll forward. They get replaced with Sivoia QS shades and current power supplies. [1][2] On paper, that reads like a line item. In the field, it can mean opening pockets, reworking power, and deciding whether the room wants Sivoia QS, Palladiom shades, or a different daylight strategy altogether.

This is where an upgrade can either get cleaner or start drifting. If the keypad language is changing, the shade language should be decided at the same time. A beautiful Palladiom keypad next to an old shade motor that still behaves like a service liability is not a finished room. It is deferred scope wearing nice trim.

What the upgrade should build toward

HomeWorks QSX is the right landing point when the house is worth keeping

Current HomeWorks is not a collection of smart switches. Lutron's platform still runs its lighting and shade control over Clear Connect RF rather than Wi-Fi, and the system is sized for more than 50,000 square feet and up to 10,000 zones. [3] That matters in large single-family homes because lighting stability should not rise and fall with the family's streaming traffic.

QSX is also the point where the project stops being defensive. The Lutron App, current shade families, Palladiom keypads, Sivoia QS, and Ketra-ready design all assume you are on the modern platform. [2][3] If the renovation is light-touch, QSX gives the estate a reliable backbone and better dimming options. If ceilings are opening and fixture design is being reconsidered, Lutron's February 3, 2026 Intelligent Lighting announcement is the bigger clue about where luxury residential lighting is headed: fixture-level intelligence, Ketra and Orluna under one residential lighting portfolio, and re-zoning without control wire because the intelligence sits in the fixture rather than in the control wiring topology. [4]

That does not mean every Illumination upgrade needs Ketra. It means there is no reason to land a serious estate on another dead-end lighting platform just before the addressable fixture layer becomes more useful.

Crestron belongs in the conversation when the house wants one control language

Some homes want lighting to stay lighting-first. Others want the kitchen touchscreen, theater remote, and bedside controls to speak one language. Crestron Home remains a sensible overlay for that. Crestron's current Lutron integration documentation says HomeWorks QSX, HomeWorks QS, RadioRA 2, RadioRA 3, and RA2 Select can be imported into Crestron Home, with Lutron lights, shades, fans, scenes, Ketra loads, and phantom keypads from QSX available inside the Crestron environment. [6]

Current Configure Pro deployments use a CP4-R v2, MC4-R v2, or DIN-AP4-R on firmware 4.11 or later. [7] That is where Cave Group's stack becomes useful rather than decorative. A CP4-R at the control layer, Lutron HomeWorks QSX owning light and shade, Sonance audio in the shared rooms, and a Kaleidescape theater scene can live together without asking the lighting system to fake being a whole-house platform it no longer is.

The mistake is not choosing Crestron. The mistake is using Crestron to paper over lighting hardware that should have been retired first.

The network is not the lighting backbone, but it still decides how calm cutover feels

HomeWorks uses Clear Connect RF and QS wiring for control, not Wi-Fi. [3] But the modern experience around it is networked: processor activation, remote support, the Lutron App, Crestron Home, and any staff-facing or whole-house interfaces. That is why an Illumination upgrade often exposes a second problem. The house has grown around an old network that nobody trusts enough to touch during the same week as the lighting migration.

On the networking side, Ubiquiti's June 25, 2026 Network 10.5 release added Test & Confirm and automatic rollback for configuration changes. [8] That sounds like a pure IT footnote until you are cutting a live estate over from legacy processors to HomeWorks QSX, re-pairing a Crestron CP4-R, and trying not to strand the property without remote access at dinner. A proper UniFi Enterprise backbone, whether that means an EFG Fortress Gateway, Pro XG switching, or cleaned-up VLANs, does not make the lighting more beautiful. It makes the migration less fragile.

What to lock before drywall closes

The decisions that prevent the second renovation

Decide whether legacy RPM panels are being preserved or replaced before millwork and trim scopes are released. That answer affects translator hardware, power, service access, and how honest the lighting budget is. [2]
Choose the keypad family before patching begins. Palladiom, SeeTouch, HQRD, HQRA, and Sunnata do not ask the same questions of box depth, finish, and multi-location behavior. [2][5]
Build a real load schedule. Mark every circuit LED, ELV, MLV, or INC/HAL. That is how you decide where current wallbox dimming is enough and where specialty load handling is required. [5]
Audit every shade by generation, power, pocket, and interface. Sivoia QED assumptions do not carry forward to Sivoia QS. [1][2]
Decide whether the house is Lutron-only at the user level or Lutron plus Crestron. If Crestron Home is part of the job, name rooms and scenes cleanly from the start so the import does not turn into cleanup later. [6][7]
Treat the network as scope, not background. Put rollback-capable managed infrastructure in place before processor cutover, not after. [8]

The upgrade standard that ages well

An Illumination upgrade is not interesting because the old system is old. It is interesting because large houses accumulate workarounds. One dimmer got swapped. One panel was never relabeled. One shade stopped reporting. One app lasted longer than it should have. If you only replace the failed piece, the renovation inherits the same uncertainty.

At Cave Group, the better standard is simple: preserve what still makes engineering sense, replace what will keep costing the owner confidence, and land the estate on a platform that can still accept good lighting design five or ten years from now. In practice, that usually means HomeWorks QSX at the lighting core, Palladiom or SeeTouch where the walls matter, current Sivoia QS or Palladiom shades where daylight is part of the architecture, and Crestron Home only when the house truly wants one control language across light, audio, video, and security. That is the kind of migration Cave Group is set up to do. It is not the cheapest path through a legacy system. It is the path that keeps the next renovation from starting in the same rack.

Lutron HomeWorks Illumination Upgrade Guide