Energy AIEdge decisions · forecasting · optimizationWe collect lighting, HVAC and nutrient-system data in a vertical farm to cut energy use while keeping growing conditions on target.
Smart Farm · Energy SavingENERGY · DIGITAL TWIN · EDGE AI · AI INFRASTRUCTURE
From energy to AI infrastructure,
run on data.
Data from the field. Validation in the virtual. Optimization in the real world.
Encall AI turns field data into an operating edge. We offer three solutions — energy management, digital twins and integrated AI infrastructure management. Adopt only the one you need, or connect them to fit your operations.
EDGE IoT NETWORK◆
ENERGY OPTIMIZATION◆
DIGITAL TWIN◆
PREDICTIVE MODELING◆
GPU CLUSTER ORCHESTRATION◆
AI WORK ASSISTANT◆
EDGE IoT NETWORK◆
ENERGY OPTIMIZATION◆
DIGITAL TWIN◆
PREDICTIVE MODELING◆
GPU CLUSTER ORCHESTRATION◆
AI WORK ASSISTANT◆
Solutions
Three solutions,
Three solutions,
each in its own field
Each solution is adopted and run independently. Data and features can be connected across them when needed.
01
Energy Management
Data-driven energy optimization
Intelligent IoT edge networks for smart farms, factories and buildings secure a reliable data source and run equipment automatically.
Learn more → 02 eAI TWINZDigital twins for large facilities
A replica of the real plant and predictive models form a virtual plant that produces optimized operating guidance, external conditions included.
Learn more → 03 eAI ForgeIntegrated AI infrastructure management
Cluster-level management of high-performance GPU resources across heterogeneous architectures.
Learn more →Each solution can be adopted on its own, with data and features connected when needed.
Energy automation
that starts at the edge
We build intelligent IoT edge networks on site to secure an uninterrupted data source. Data is analyzed right where it is collected, and heating, cooling, lighting, HVAC and nutrient systems control themselves.
- Site-specific edge networks — sensors and gateways designed for each smart farm, factory and building
- A reliable data source — equipment, environment and power data, standardized and kept as an asset
- Automated operation — real-time control and anomaly detection with rules and AI models
- eAI Desk, AI for energy — ask the accumulated energy data and get savings reports right away
ARCHITECTUREPick a site, then hover an item to follow its flow
01 · FIELD
IoT edge network
02 · DATA
04 · ACTION
DEMO · illustrative values
- 01Field sensing
Sensors and equipment across farms, factories, buildings and renewables/ESS
- 02Data collection
Power, environment, equipment and generation data standardized at the edge
- 03Edge AI
Decisions on site, plus forecasting and optimization eAI Desk, AI for energy
- 04Automatic control
HVAC, lighting, nutrient and ESS control, peak shaving, savings reports
AI-based EMS design for the Gapado RE100 village project. Wind and solar generation and island-wide consumption data are gathered at the edge and run together.
Island MicrogridSolar PV and ESS with a charge/discharge algorithm cut demand in peak-tariff hours, lowering both energy use and electricity bills.
PV · ESS · Peak ShavingWe build the plant twice.
Virtually, first.
We build a digital replica of the real plant and a virtual plant driven by data-based predictive models. After simulating with external conditions, we derive optimized operating guidance for energy savings and efficiency.
8Chillers twinned live1 turbo · 7 screw
5Cooling towers & pumps linked3 chilled- · 5 condenser-water pumps
500+Measurement tagsLive MQTT collection
32KPI items7 categories
13Chart typesheatmaps, 3D scatter and more
Live 3D monitoring
A 3D plant that mirrors live measurements. Select the chillers and the building turns translucent, revealing eight chillers, piping and operating/alarm badges.
Equipment-level digital twin
Beyond the whole plant, each chiller has its own twin. Open a unit in section view and tap the compressor, evaporator or condenser to read temperature, pressure and current.
Real and virtual on one screen
Split the view side by side, top and bottom or three ways, or overlay a lens, to compare today's plant with the simulated one from the same angle — even under different weather, sunny or rainy.
A virtual plant that never stops
The whole system — eight chillers, chilled- and condenser-water pumps, brine tanks and five cooling towers — runs as a virtual plant. AI servers simulate 6 hours to 7 days nonstop, comparing efficiency trends, power mix and operating history.
Measurement data analysis
Pick from 500+ measurement tags and view them in 13 chart types such as lines, heatmaps and 3D scatter, with cross-equipment correlation and CSV/Excel export.
Images: eAI TWINZ live screens · CKD BiO Ansan Plant
- 01
Replicate
A replica of the real plant: equipment, piping and sensors as they are
- 02
Predict
A virtual plant driven by predictive models trained on operating data
- 03
Simulate
Scenario checks with external factors such as temperature, humidity and load
- 04
Optimize
Optimized operating guidance for energy savings and efficiency
CHILLER PLANT · POWER PROFILE
ILLUSTRATIVE
Hover the chart to see values and expected savings by time of day · illustrative data
Encall AI
×
VEOLIA
FLAGSHIP CASE · Live
CKD BiO Ansan Plant
Integrated chiller control for energy savings. We replicate the refrigeration systems of the Ansan plant, operated by Veolia, as a digital twin and optimize energy use with model-based integrated control.
NEXT · In progress
- Ourhome Gyeryong Plant
- Veolia-managed facilities in Korea
Different architectures,
one cluster
GPU servers of different architectures are grouped into clusters and managed as one resource pool. Researchers submit jobs on a portal, and eAI Forge places them on the best-fitting nodes.
- Heterogeneous architectures — x86 GPU servers and ARM-based DGX Spark in a single pool
- Cluster-level resource management — node and GPU allocation, queue and quota policies, usage visibility
- Web portal + SSO — pick a node, environment and resources, and JupyterLab or a desktop opens right away
Reference · Chung-Ang University AURORA cluster
4Nodes
6GPU
296CPU cores
1.2TBMemory
Chung-Ang University AURORA · x86 DeepGadget + ARM DGX Spark (GB10) heterogeneous setup · live screen
Products
Products that solve
Products that solve
problems in the field
One product for each of solutions 01 · 02 · 03 — what each screen does and where it runs.
Applied to solution 01 · in daily use at Encall AI
eAI Desk
An AI work assistant tailored to your domain
An AI workspace that connects field data with business knowledge and adapts to each domain. It runs as an energy-specialized assistant in solution 01 and can be tailored to other domains the same way.
- Energy Applied
- Internal work Live
- Manufacturing
- Facilities
- Research · Education
- More domains
- Domain tailoring
- A model per task
- Reports & proposals
- Notes & knowledge search
The chat on this page is a site guide.
CKD BiO Ansan Plant · with VEOLIA
eAI TWINZ
Digital-twin operation of large facilities
Integrated chiller control for energy savings. A replica of the real plant and predictive models produce optimized operating guidance.
- Equipment status at a glance
- Predictive models & scenarios
- Operating guidance
Chung-Ang University · AURORA cluster
eAI Forge
Heterogeneous GPU cluster management
Unifies GPU servers of different architectures into clusters, with a web portal from job submission to monitoring.
- Node & GPU status
- Job submission & tracking
- Cluster usage management
Image: Chung-Ang University AURORA · eAI Forge live screen
References
Running today,
Running today,
in the field
Loading map…
-
HQ
Encall AI
18, Dongil-ro 34-gil, Gwangjin-gu, Seoul
eAI Desk in daily use - Live CKD BiO Ansan Plant eAI TWINZ · integrated chiller control for energy savings · with VEOLIA
- In progress Ourhome Gyeryong Plant eAI TWINZ · digital-twin facility management
- Reference Gapado Island, Jeju Energy optimization · AI-based EMS for an RE100 village
- Live Chung-Ang University AURORA eAI Forge · heterogeneous GPU cluster
Sites are shown at city level; sites that cannot be disclosed are left off the map · Map © OpenStreetMap contributors, OpenFreeMap
Turn field data
into an operating edge.
Energy optimization, digital twins, AI infrastructure — we'll help you decide where to start.