Building management system replacement: replace, keep or overlay
Milvian AI teamFacilities and infrastructure intelligence
Replace controllers when they are unsupported or failing, not because the front end is weak. Where the controllers still work, keep them and add a layer above that fixes visibility.
Why replacement decisions go wrong
BMS replacement is usually triggered by frustration with the front end: slow graphics, poor trends, one contractor who holds the passwords. A full replacement fixes those, but it also swaps controllers that may have years of life left, and it can cost months of disruption.
The other mistake is waiting too long. Controllers that are out of support fail without spares, and an unsupported operating system on the front end is a security problem.
Four situations
| Situation | Better path | Why |
|---|---|---|
| Controllers unsupported or failing. | Replace those controllers. | Spares and security updates are gone. |
| Controllers fine, front end weak. | Keep and overlay. | Visibility can be fixed without touching control. |
| Proprietary lock-in. | Overlay now, replace in phases. | Open data first, then move controllers as they age. |
| Major refurbishment. | Replace with open protocols. | Walls and ceilings are open anyway. |
What a replacement must deliver
| Requirement | Why it matters here | Ask the vendor |
|---|---|---|
| History carried over. | Trend history is how you prove later savings. | How do you migrate existing trends? |
| Phased cutover. | Buildings cannot lose control during work. | What is the cutover plan per system? |
| Open protocols and rights. | The next change should be cheaper. | Who holds programming rights after handover? |
What makes now a good time
Building performance standards such as New York's Local Law 97, which tightens limits in 2030, and the recast EU EPBD, which targets the worst 16 percent of non-residential buildings by 2030, reward buildings that can show measured performance. Controls matter here too: LBNL found 40 percent of air handlers reporting a fault on any given day across 60,000 pieces of equipment, and a new front end only helps if those faults get fixed.
How to run it
- Audit controllersRecord support status, spares and firmware for each.
- Overlay firstRead existing points into a new layer to secure history and visibility.
- Replace in risk orderStart with failing or unsupported controllers.
- Verify each phaseCompare energy and comfort before and after against a baseline.
Where Aqueduct fits
Aqueduct supports all three paths. It reads existing controllers to keep history and add a portfolio view, replaces failing equipment control through Smart Controls over BACnet and Modbus where control is enabled, with owner approval, and can be specified for new builds. Weather-normalized baselines in aQ EnergySmart show whether each phase delivered.
Sources
| Source | Link |
|---|---|
| NYC Department of Buildings, Local Law 97 | nyc.gov |
| European Commission, Energy Performance of Buildings Directive | energy.ec.europa.eu |
| LBNL, Crowe and Granderson et al., 2023 | bies.lbl.gov |
