- Johnson Controls
- India's first on-premises AI-powered chiller plant optimisation in healthcare
India's first on-premises AI-powered chiller plant optimisation in healthcare
Impact Highlights
How Johnson Controls and Graylinx deployed the Chiller Plant Optimizer at Jupiter Hospital, Dombivli, redefining what intelligent building technology can achieve in critical clinical infrastructure.
~0.75 kW/TR plant efficiency achieved
Up to 500 kWh daily energy savings
1.2–1.5 lakh kWh/yr estimated annual savings
Jupiter Hospital, Dombivli
The Customer
Jupiter Hospital, Dombivli is a state-of-the-art green multi-specialty hospital on Kalyan-Shil Road, serving the Dombivli and Kalyan regions of the Mumbai Metropolitan Region. Part of the Jupiter Lifeline Hospitals group, the facility has a planned capacity of 500 beds, including ICU/PICU/NICU 105 beds and 19 modular operation theatres, and delivers advanced critical-care and quaternary medical services across more than 30 specialties.
|
GROUP Jupiter Lifeline Hospitals |
LOCATION Dombivli, Mumbai Metropolitan Region |
|
CAPACITY 500 beds | ICU/PICU/NICU 105 beds | 19 modular OTs |
DEPLOYMENT Jan-Mar 2026 | Go-live Feb 2026 |
|
SPECIALTIES 30+ clinical specialties |
BUILDING TYPE IGBC Green-certified hospital |
THE CHALLENGE
The twin demands of comfort and cost efficiency
Hospital HVAC systems present a unique optimization challenge: they must deliver uninterrupted temperature and humidity control across critical clinical environments, while doing so at the lowest possible cost of energy. For a facility of this scale, these two demands are not easily reconciled.
Before the CPO deployment, the chiller plant at Jupiter Hospital was operating in manual mode, with plant efficiency (kW per tonne of refrigeration, or kW/TR) exceeding 0.80. The absence of automated, dynamic control meant that stage-up and stage-down sequencing was handled manually, introducing operational variability and leaving meaningful efficiency gains unrealised.
The hospital's core requirements were:
- Fully automated plant room operation, eliminating the need for manual intervention
- Continuous, real-time optimisation aligned with variable load and ambient conditions
- Lift minimisation to achieve the best plant COP and the lowest possible kW/TR
- High plant availability to meet the non-negotiable uptime requirements of critical healthcare infrastructure
The Solution
Chiller Plant Optimizer (CPO)
The Chiller Plant Optimizer is an on-premises, AI-powered solution that delivers continuous, real-time adaptive control of a central chiller plant. Developed in partnership with Graylinx, a specialist in AI-based building energy optimisation and part of the NR Group of Companies, the CPO integrates natively with Johnson Controls' chiller performance expertise.
This is the first on-premises, non-cloud, AI-powered Central Plant Optimization deployment in India's healthcare sector. The plant is a 2 x 350 TR water-cooled chiller configuration in a primary-secondary arrangement serving two secondary zones.

Key capabilities deployed:
- Real-time adaptive control: the AI continuously analyses plant conditions, load, and ambient data to make dynamic control decisions, with no manual triggers required
- Dynamic set-point optimisation: set points for chilled water supply temperature, condenser water temperature, and other key variables are automatically adjusted to maximise efficiency at every operating point
- Lift minimisation: the CPO optimises the differential between evaporator and condenser temperatures to ensure the plant always operates at its best Coefficient of Performance
- Automated stage sequencing: clean, logic-driven stage-up and stage-down of chiller units replaces manual sequencing, improving operational consistency and reducing wear
- Native CPO decision feedback: the system provides continuous feedback on its own performance decisions, enabling ongoing learning and refinement
Why Johnson Controls
Three capabilities that defined this project's success
Performance improvement
From manual operation to AI-powered precision
The CPO deployment improved plant energy intensity from above 0.80 kW/TR in manual mode to a sustained 0.75 kW/TR under AI control. A meaningful gain on a plant that was already well-engineered.
|
BEFORE | Manual Mode 0.80+ kW / TR Variable, operator-dependent sequencing. No dynamic optimisation of lift or set-point. |
AFTER | CPO Active 0.75 kW / TR Continuous AI-driven lift minimisation, dynamic set-point control, and automated stage sequencing. 24/7. |
Up to 10% improvement in kW/TR | Equivalent to removing approximately 150 kWh of waste per operating day
Energy savings
Measured, verified, and continuing to compound
Savings are not projected estimates. They are measured outcomes from the live plant. The CPO's decision-table feedback loop means performance improves as the AI learns the plant's behaviour over time.
|
DAILY SAVINGS (PEAK) 500 kWh per day Measured under CPO control. Sustained range: 300 to 500 kWh per day across varying load conditions. |
ANNUAL SAVINGS (ESTIMATED) 1.5 lakh kWh per year Translates to meaningful carbon reduction and direct operating cost savings, year after year, without manual intervention. |
Sustainability alignment
This deployment supports IGBC green-building certification points and aligns with India's national push toward decarbonisation of critical buildings, directly contributing to Jupiter Hospital's green-building standing.
Outcomes - What changed from day one
Energy intensity: 0.80+ to 0.75 kW/TR
A step-change in plant efficiency was achieved on day one, with further improvement expected as the AI model learns the plant's seasonal and load-dependent behaviour.
Zero manual intervention in plant sequencing
Stage-up and stage-down of chiller units is now handled entirely by the CPO, eliminating operator variability and improving consistency across all operating hours.
Hospital data stays on-premises
No operational data is routed to external servers. The CPO runs on local infrastructure, meeting the data security requirements of a critical-care facility.
IGBC green-building certification support
The CPO's measurable energy performance directly supports Jupiter Hospital's IGBC rating, aligning with India's growing emphasis on sustainable healthcare infrastructure.
India's first on-premises AI-powered CPO deployment in a hospital
This project establishes a clear proof of concept for non-cloud AI in critical healthcare infrastructure, where data security, uptime, and precision of control are non-negotiable. It demonstrates that the efficiency ceiling for well-designed chiller plants is not fixed.
“As per our expectations for best-in-class performance, the JCI premium, energy-efficient water-cooled chillers are delivering excellent results. Additionally, the CPO system is operating effectively, ensuring optimal power consumption and supporting our goal of achieving green building platinum certification.”
Neelesh Shinde
Group CTO & Head - Projects
Conclusion
A new standard for intelligent buildings in Indian healthcare
The Chiller Plant Optimizer deployment at Jupiter Hospital, Dombivli is a milestone in the application of building intelligence to healthcare infrastructure in India. By combining Johnson Controls' deep HVAC application expertise with Graylinx's on-premises AI capability, the project delivers a solution that is secure, precise, and built for the operational realities of a critical-care environment.
It reflects Johnson Controls' commitment to outcomes over deliverables: not simply installing technology, but taking responsibility for the results it produces. As India's healthcare sector grows and its green-building ambitions deepen, this project sets a benchmark for what intelligent building systems can achieve in the country's most demanding operational environments.

















.jpg?la=en-IN&h=320&w=720&hash=FC6020117990874E3CCEA5C9EDFFF0AB)














.jpg?la=en-IN&h=310&w=720&hash=CB9077B5CBA22B0C0A1F2C6EA81CC63A)


.jpg?la=en-IN&h=320&w=719&hash=7D4D720F7C57C8374A83F8A70533B32F)
.jpg?la=en-IN&h=306&w=720&hash=50F8B4CB07BD729D1C73C3949F62A93F)

.png?la=en-IN&h=320&w=720&hash=E80AA15C270907356AEB3B8427334ED9)


