ClimaMind Energy IntelligenceClimaMind
Book a Demo
Field evidence

Audited savings across hospitality, commercial towers, and critical infrastructure

Each deployment pairs reinforcement learning with layered safety, verified against alternating-day or contractual baselines. These case studies show how ClimaMind supports hospitality portfolios, mixed-use developments, transportation hubs, industrial plants, civic venues, and data centers without disrupting operations.

20 verified deployments spanning hospitality, commercial & mixed-use towers, transportation hubs, healthcare campuses, industrial manufacturing, public venues, infrastructure energy hubs, and data centers
Savings validated through alternating-day tests or contractual guarantees where available
Layered safety and manual override kept intact for every site

Hospitality & Resorts

Hotels and resorts maintaining 24/7 guest comfort while deploying reinforcement learning across plant operations.

Shanghai, China

Deployment snapshot

Agile Marriott Marquis Shanghai

39-story hospitality tower running 24/7 introduced AI orchestration across chilled water and domestic hot water systems.

Key metrics

Building area
45,000 sqm
Guest rooms
720
System scope
Cooling plant + hot water plant

Project background

  • Commissioned in 2015 with continuous 24/7 operations.
  • Chiller plant supplies cooling while boilers and air-source heat pumps support heating and domestic hot water.

Optimization approach

  • AI-driven global optimization governs chilled water setpoints under normal cooling conditions.
  • Predictive load balancing stabilizes the hot water system by coordinating heat supply and demand.

Results

  • Dual-system AI control keeps operating conditions near peak efficiency for both comfort and domestic hot water loads.

Lingang, Shanghai, China

Deployment snapshot

World Laureates Forum Venue Hotel

Permanent venue hotel for the World Laureates Forum deployed modular RL across a parallel chiller plant with tight comfort limits.

Key metrics

Verified savings
18.5%
Deployment window
28 days (installation + tuning)
Plant assets
4 chillers / 5 CHW pumps / 5 CW pumps / 8 cooling towers

Project background

  • Central plant configured in full parallel with magnetic-bearing chillers and Grundfos pump sets.
  • Comfort guardrails and event schedules left narrow bands for frequency adjustments.

Optimization approach

  • On-site RL deployment enforced pump frequency bands between 40-50 Hz and tower frequency between 36-50 Hz.
  • Tuning ran inside a secure on-premise environment to protect venue operations.

Results

  • Commissioning and optimization completed within 28 days while maintaining comfort limits.
  • Verified 18.5% energy savings after stabilization of AI-driven sequences.

Commercial & Mixed-Use Towers

Landmark office and retail campuses balancing premium tenants, tariff windows, and shared energy plants with reinforcement learning.

Shanghai, China

Deployment snapshot

West Bund International AI Center

Shared cold plant for a 200m tower and retail podium combines RL-based runtime optimization with tariff-driven ice storage strategies.

Key metrics

Verified savings
21.6%
Cold plant assets
5 chillers / dual-loop pumps / ice storage
Mixed-use area
80,000 sqm office & retail

Project background

  • Prime office tower and mall share chilled water, including dual-mode chillers and ice tanks.
  • Office and retail loads fluctuate heavily between dayparts and seasons.
  • Tariff windows require coordinated storage dispatch to avoid peak penalties.

Optimization approach

  • Runtime optimization ensures chilled water plants run at global optimum in mechanical modes.
  • Load forecasting and tariff-aware scheduling determine ice charge/discharge windows.
  • Unified dashboards align operations and finance on consumption vs. cost trade-offs.

Results

  • 21.6% verified energy savings after joint runtime and tariff optimization.
  • Delivered repeatable template for campuses with thermal storage assets.
  • Improved transparency for operators through aligned cost-performance metrics.

Shanghai, China

Deployment snapshot

Shanghai IFC Towers & Mall

Twin LEED Gold towers, a Ritz-Carlton hotel, and IFC Mall share an integrated plant requiring cross-program load management.

Key metrics

Verified savings
35%+
Gross floor area
400,000 sqm
Program mix
Offices / luxury hotel / club / retail

Project background

  • South tower rises 260 m and north tower 250 m within Lujiazui's core business district.
  • Tenants span global financial institutions alongside a Ritz-Carlton hotel and branded clubhouse.

Optimization focus

  • Mixed-use loads require unified control sequences for office baseloads and hotel comfort KPIs.
  • Project planned while operations continue across tier-one retail and hospitality venues.

Results

  • Verified energy savings exceed 35%, bringing audit-ready proof for stakeholders across asset classes.

Transportation Hubs

Airports and rail hubs safeguarding mission-critical passenger comfort under volatile load patterns.

Shanghai, China

Deployment snapshot

Shanghai Pudong Airport Satellite Hall

Digitized a 622,000 sqm terminal with direct chilled water supply by building hydraulic and temperature-field models that feed MPC controls.

Key metrics

Verified savings
22.6%
Commissioning
Completed 2021
System profile
Triple-pump direct supply / full-VFD AHUs

Project background

  • World's largest satellite terminal with S1/S2 concourses and comfort-critical zones.
  • Direct chilled water delivery after removing HX bottlenecks; all major equipment on VFD.
  • Chronic high flow and low delta T undermined energy efficiency.

Optimization approach

  • Model predictive control to coordinate chillers, pumps, and airside loads.
  • Hydraulic and temperature-field digital models maintain visibility for operators.
  • Parameter libraries exposed for visualization, audit, and fast re-tuning.

Results

  • Initial rollout exceeded 15% savings acceptance threshold.
  • Long-term alternating-day assessments confirmed 22.6% savings.
  • Resolved high-flow/low-delta-T issue while maintaining passenger comfort envelopes.

Shantou, China

Deployment snapshot

Shantou High-Speed Rail Station

Mission-critical rail hub now coordinates chillers, pumps, towers, AHUs, and PAUs through a theory-of-mind RL agent.

Key metrics

Verified savings
25%+
Station footprint
330,000 sqm / 11 platforms / 24 tracks
Cold plant
4 variable-speed centrifugal chillers

Project background

  • Largest transportation hub in Eastern Guangdong with uneven passenger flows.
  • Cold plant built on four VSD centrifugal chillers plus balanced pump/tower sets.
  • Manual tuning could not keep pace with load swings or comfort guardrails.

Optimization approach

  • Theory-of-mind transfer RL agent governs chillers, pumps, towers, AHUs, and PAUs.
  • On-site commissioning validated control signal delivery and override pathways.
  • Safety envelopes keep manual fallback available during all optimization cycles.

Results

  • Verified energy savings above 25% during acceptance testing.
  • Operators reduced manual interventions, improving stability and uptime.
  • Extended equipment life with smoother staging and frequency control.

Healthcare Networks

Hospital campuses coordinating critical HVAC assets with layered safety for clinical reliability.

Shanghai, China

Deployment snapshot

Yangpu District Central Hospital

Medical facility modernization program leverages ClimaMind RL controls to protect patient comfort while reducing central plant energy use.

Key metrics

Verified savings
25.2%

Project background

  • Hospital operations require strict thermal stability across critical care areas.

Results

  • Reported 25.2% energy savings validated through project acceptance testing.

Xiamen, China

Deployment snapshot

Xiamen Xiang'an Hospital

Coastal hospital adopted ClimaMind RL strategies to deliver significant savings while maintaining continuous clinical operations.

Key metrics

Verified savings
33.2%

Project background

  • Large-scale healthcare campus with stringent comfort and reliability requirements.

Results

  • Project accepted with a verified 33.2% reduction in energy consumption.

Industrial Manufacturing

Process cooling systems in pharmaceutical and PV plants balancing strict production constraints with energy savings.

Suzhou, China

Deployment snapshot

GSK Suzhou Pharmaceutical Plant

Pharmaceutical manufacturing site implemented ClimaMind AI controls to optimize HVAC energy use across production spaces.

Key metrics

Verified savings
12.1%

Project background

  • Production environment demands tight control over temperature and humidity to safeguard product quality.

Results

  • Energy audits confirmed a 12.1% reduction in plant energy consumption.

China

Deployment snapshot

Zhonghuan Low-Carbon PV Plant

Photovoltaic manufacturing facility uses ClimaMind RL orchestration to contain energy costs across process cooling assets.

Key metrics

Verified savings
18.1%

Project background

  • Industrial site with continuous cooling demand for photovoltaic production lines.

Results

  • Verified 18.1% savings on energy consumption following optimization rollout.

Public & Exhibition Venues

Civic and exhibition venues running year-round climate control while reducing consumption.

Chongming, Shanghai, China

Deployment snapshot

Chongming Flower Expo Venues

Expo campus venues on Chongming Island deploy ClimaMind RL optimization to balance exhibition comfort with verified energy reductions.

Key metrics

Verified savings
25.1%

Project background

  • Flower Expo halls require continuous climate control for year-round exhibitions.

Results

  • Alternating-day comparison reports a 25.1% reduction in energy consumption.

Infrastructure & Energy Hubs

District energy centers dispatching multi-building loads with reinforcement learning and storage coordination.

Shanghai, China

Deployment snapshot

G60 Rafael Sky City Energy Center

23 high-rise towers share an ice storage plant that now balances comfort, grid pricing, and energy headroom with reinforcement learning.

Key metrics

Verified savings
10.12%
Estimated monthly power cost reduction
CNY 150K
Plant assets
7 chillers / 12 cooling towers

Project background

  • Mixed heat and cooling demand across a million-square-meter commercial campus.
  • Ice storage charged after 22:30, daytime discharge tied to peak tariffs.
  • Legacy strategies could not jointly optimize energy and tariff windows.

Optimization approach

  • Modular EcoWisoul RL agents with load forecasting driven by weather and TOU tariffs.
  • Coordinated staging for chillers, cooling towers, and chilled/condensing pumps.
  • Monthly reporting aligned with utility sub-meter validation and city science grants.

Results

  • Savings visible within first comparable-day test window.
  • Seasonal monitoring shows 32% winter and 10.3% summer reductions.
  • Dual optimization on kWh and tariff cost baseline delivered.

Shanghai, China

Deployment snapshot

World Expo Energy Station

District energy station from the World Expo introduced ClimaMind RL to coordinate central plant assets and distribution loops.

Key metrics

Verified savings
18.9%

Project background

  • District-scale plant supplies multiple connected venues with cooling and heating.

Results

  • Post-deployment measurement reports an 18.9% reduction in energy consumption.

Data Centers

Carrier and AI data centers pushing PUE lower across mechanical, free-cooling, and hybrid modes.

Shanghai, China

Deployment snapshot

China Unicom Zhoupu Data Center

Four-story carrier facility met a 19% savings guarantee while running low IT load conditions with magnetic-bearing chillers.

Key metrics

Verified savings
19%
Cooling plant
3 x 3340 kW magnetic VSD chillers
Recognition
Featured at National DC Cooling Forum

Project background

  • 21,500 sqm total area with contract guarantee requiring >=19% energy reduction.
  • Low IT load scenarios limited traditional savings headroom.
  • Chilled plant built on three identical magnetic-bearing centrifugal units.

Optimization approach

  • EcoWisoul modular RL agents orchestrated chillers, chilled water, and condenser loops.
  • Dynamic policy tuning preserved stability during very low load periods.
  • Interactive savings dashboard exposed KPIs to both operators and executives.

Results

  • Delivered 19% verified savings and cleared acceptance testing.
  • Project showcased at the 7th National DC Cooling Energy Forum.
  • Set baseline for replicating RL optimization across carrier cloud sites.

Beijing, China

Deployment snapshot

21Vianet Beijing Tengren DC

High-performing colocation site added RL optimization to free-cooling, pre-cooling, and mechanical modes to push PUE gains further.

Key metrics

Free cooling savings
17.23%
Mechanical cooling savings
7.54%
Cooling assets
4 x 2602 kW VSD chillers + HX + tower sets

Project background

  • Twelve data halls (4 large, 8 small) with seasonal PUE between 1.10 and 1.36.
  • Cooling loop fully serial, leaving limited room for traditional retrofits.
  • Operations already run by a dedicated professional team with frequent setpoint tuning.

Optimization approach

  • RL policies targeted cooling-side COP across free, pre-cooling, and mechanical modes.
  • Chiller discharge, pump frequency, pump staging, and tower speed treated as unified action space.
  • Automatic mode switching maintained safety checks for every transition.

Results

  • 17.23% savings in free-cooling mode and 7.54% in mechanical mode.
  • Operators retained high visibility and control despite automation depth.
  • Template extended to other north China colocation campuses.