One building.
One intelligence.
Local adaptive control for energy systems.
ORLA learns how real installations respond and evaluates actions within operator-defined limits. The same local approach can support a single installation or a coordinated energy portfolio.
The intelligence
is in the interaction.
A modern building contains many smart systems. But the best decision for each device in isolation is not necessarily the best decision for the building.
Kitchen → HVAC
Heat from breakfast preparation changes the actual heating demand of nearby spaces.
PV → Hot water + BESS
Surplus generation can prepare hot water or charge a battery — depending on what will be needed next.
EV → Building power
EV charging can coincide with HVAC start-up. Coordinating loads can avoid a shared peak.
How should the whole building operate, considering what happens next?
Sense now. Predict next. Decide together.
ORLA combines operational data, forecasts and operator constraints. In the local pilot setup, ORLA and Home Assistant run in separate containers. Home Assistant connects to available data and authorised control points.
Sense
Zone temperatures, occupancy, equipment, energy, weather, prices, PV generation and battery SOC.
Predict
Future zone needs, temperature changes, power peaks and the cost of preparing the building.
Decide
Which resources to use, when and how intensely. Evaluate consequences across the whole system.
Coordinate
An approved strategy can act within comfort, process and equipment limits. Operational feedback informs future learning.
You define the boundaries. ORLA chooses the strategy.
A site can set goals for comfort, cost, demand and flexibility. The available equipment, operating limits and validation determine which goals ORLA can pursue.
COMFORT
Prepare spaces before they are needed.
SAVINGS
Reduce the cost of the whole system.
PEAK
Coordinate demand ahead of the peak.
BATTERY CARE
Use building flexibility before the battery when it makes sense.
SELF-CONSUMPTION
Give local generation its greatest value.
FLEXIBILITY
Offer only a demand change the building can deliver.
Tested on real operational data.
Two separate HVAC studies used historical BMS data from a real building.
Thermal prediction
0.23°C MAE1-hour temperature forecast vs 0.27°C persistence baseline.
13 months of BMS archives, four rooms and four held-out monthly tests covering different seasons.
Control strategy evaluation
24–33% lowerModelled heating-demand indicator vs historical BMS control.
119.5 days, over 34,000 evaluated decisions and four HVAC zones. The indicator is based on valve operation. Metered kWh and bill savings require active M&V.
How the studies were evaluated
The persistence forecast keeps the current temperature unchanged for one hour. MAE is the average absolute difference between predicted and observed temperatures. The retrospective control study compares a candidate policy with historical BMS behaviour using a valve-based indicator.
From individual assets to energy portfolios.
Choose where the intelligence runs and how many assets to coordinate. Each step beyond the current edge pilot has its own validation requirements.
Edge / on-prem
Local models, learning and decisions can run on site without a continuous cloud connection. Integration and control permissions are qualified for each site.
Cloud
Optional portfolio coordination, analytics, reporting and flexibility aggregation. Execution stays within each site’s local limits.
Embedded / MCU
Selected functions could run directly on OEM devices, subject to processor, memory and integration limits. Full ORLA currently targets edge hardware.
Device
Learn the response of a controllable device.
Installation
Evaluate actions within process limits.
Building / site
Coordinate qualified installations against a shared goal.
Portfolio
Aggregate the flexibility offered by participating sites.
Microgrid / VPP
Coordinate a portfolio response. Sites retain local execution.
Cloud coordination. Local control.
A central platform can send an energy objective without direct access to every device. Sites can return the flexibility and telemetry needed to coordinate a response.
Cloud / EMS / VPP
Sets objectives, coordinates the portfolio, aggregates flexibility and connects to market processes.
Local ORLA / edge
Models the process, checks operating limits and executes authorised actions. Operational data can stay on site.
Existing BMS / OEM automation
Retains baseline control, independent equipment protection, safety limits and fallback procedures.
This architecture can reduce remote exposure. Each deployment needs authentication, network segmentation, secure updates, decision logging, withdrawal procedures and security testing. Local operation during a connectivity loss depends on site-tested scenarios.
One winter morning. Hundreds of connected decisions.
Explore how a hotel’s context changes. This is an illustrative scenario, not a record from a deployment.
The building wakes up
It is cold outside. The kitchen will start shortly.
Comfort / predictionHVAC starts heating according to its schedule.
ORLA prepares the required zones while accounting for heat gains that will arrive later.
Starting with HVAC. Expanding toward multi-asset coordination.
HVAC is the current focus of validation. Thermal inertia can help shift demand, while batteries can provide a faster response. Coordinating them with other assets is a development direction that requires site-specific validation.
How much of the response could the building deliver?
A building can contribute part of a demand reduction, leaving the battery to supply the balance. Move the slider to explore the split.
Illustrative allocation for a 100 kW reduction. Effects on battery throughput and lifetime require measurement and verification (M&V).
How much demand can the site change, and when?
A flexibility envelope describes feasible changes against a baseline, given the forecast and local constraints. An EMS, VPP or aggregator could use it to coordinate a response. Delivery must be validated on site.
Learn the site. Validate decisions. Enable bounded control.
The five stages lead from data access to controlled activation. Agree on a baseline, stop criteria and return to baseline control before active M&V.
Connect data
Link BMS, meters, controllers and sensors.
Learn
ORLA learns the dynamics of the actual site.
Shadow mode
Evaluate decisions without affecting equipment.
Bounded autonomy
Enable selected resources after validation.
Expand
Add resources and optimisation objectives.
Every installation is different. Deployment should be repeatable.
ORLA aims to reduce site-specific modelling effort. Point mapping, integration, testing and commissioning still need to be completed at each site.
Modelling-intensive workflow
- Analyse the installation
- Build and calibrate a site model
- Tune the control strategy
- Repeat engineering at the next site
ORLA workflow
- Integrate operational data
- Define objectives and constraints
- Learn the installation response
- Validate candidates and adapt
For integrators, OEMs and portfolio operators, the pilot measures time to integrate, engineering hours, reuse and maintenance cost. A proposed renewable partner licence can support repeat deployments within an agreed scope.
Keep the infrastructure. Add the intelligence.
ORLA complements existing BMS, EMS, local controllers and OEM systems. The integration scope is matched to the equipment and interfaces at the site.
Local decisions.
Optional cloud.
Sensing, prediction, decisions, control and fallback can run inside the building. Reporting, portfolio monitoring and external integrations can use optional services.
Where to start and where to expand.
HVAC is the validation focus. The applications below illustrate potential uses and require qualification for each site.
Buildings and portfolios
Cost, comfort, power peaks and PV/BESS coordination.
Hotels
Guest comfort, HVAC, hot water and occupancy forecasts.
Retail and refrigeration
Repeatable sites and processes with their own inertia.
EMS, integrators, OEMs
Autonomous intelligence above existing automation.
Start with a pilot.
Start with a free assessment of your data, equipment and objectives. A pilot can progress from site qualification and shadow to controlled operation and active M&V against an agreed baseline.
Building owner or operator
Assess your buildingHVAC / BMS integrator
Become a deployment partnerEMS / VPP operator
Explore flexibility integrationOEM / technology partner
Discuss technology integrationPilot scope and success criteria are agreed for each site.