Altius Surya Tech Pvt Ltd
T 172, Concorde Silicon Valley, Neeladri Road, Electronic City Phase 1, Bangalore, Karnataka, India 560100
PLATFORM LIVE

AI-Kit™ Digital Twin

Real-Time AI Optimization Platform for Solar Pumping Systems
Digital Twin Machine Learning Explainable AI v4.0.0
Integrated Renewable-Energy Intelligence

One AI platform for solar pumping, PV performance and solar-resource simulation

Monitor operating conditions, understand AI decisions, compare control strategies and launch specialised solar applications from a single client-facing platform.

Platform status Operational
Digital Twin Live
AI controller RF-AI Ready
Operating modes 3
Release Executive Edition
AI-Kit Product Suite

Integrated Solar Intelligence Platforms

Each application addresses a distinct operational need while remaining part of the Altius AI-Kit ecosystem.

LIVE APPLICATION

AI-Kit Digital Twin

Real-time solar-pump simulation, VFD telemetry, machine-learning recommendations and operational validation.

  • Live pump and VFD operating state
  • Baseline, Rule-AI and RF-AI control
  • Explainable frequency recommendation
Open Digital Twin
PV INTELLIGENCE

PV Cleaning Platform

Analyse module images and operating information to prioritise cleaning and understand potential soiling-related losses.

  • Image-assisted PV inspection
  • Cleaning-priority recommendations
  • Maintenance history and reports
Launch PV Cleaning
SOLAR SIMULATION

Sun Simulator

Explore irradiance, weather and location scenarios to understand expected solar-system behaviour.

  • Location-based solar scenarios
  • Weather and irradiance simulation
  • Expected-energy assessment
Launch Sun Simulator
CLOUD OPERATIONS

AI-Kit Cloud

Centralised telemetry, pilot monitoring, analytics, documents and multi-site operational visibility.

  • Fleet and pilot dashboards
  • Cloud telemetry and analytics
  • Technical resources and reports
Open AI-Kit Cloud
Virtual Plant
Weather Engine
PV Model
AI Engine
EC200U Waiting
Modbus Waiting
Cloud Connected

Digital Twin Asset Configuration

Pilot / Profile Pilot 1
Site Bangalore
PV Capacity 3.0 kWp
VFD INVT GD20-1R5G
Motor Power 1.5 kW
Motor Voltage 380 V
Motor Current 4.0 A
Pump Rated Flow 22 m³/h
Pump Rated Head 42 m
Tank Capacity 20 m³
Modbus Slave ID 1
Serial Settings 19200, 8-E-1

EC200U and Synthetic Modbus Live Status

Synthetic VFD server Checking...
EC200U link Waiting for first poll
Modbus requests 0
Last poll None
Last register None
Register count None
Function code None
Request type None

The EC200U changes to Connected only after a real Modbus request reaches the isolated DTECH adapter.

Solar Weather Input

Current source: Live location data

Live mode uses current irradiance, temperature and cloud cover. Demo mode generates artificial daytime values for exhibitions and night testing.

Selected city Bangalore
Last update Live 00:00:23
Weather Night

System State

Pump RUNNING
Control mode PV-AUTO
Fault None
Virtual device INVT GD20
Power source SOLAR_ONLY
Stop reason No usable solar power

Digital Pump Station Mimic

Solar and Weather
0 W/m²
Night, 21.2 °C
PV and VFD
0.0 kW
0 Hz
💧
Pump and Tank
0 m³/h
62.2%
Frequency
0 Hz
Motor Current
0 A
Motor Voltage
0 V
DC Bus
0 V
Power
0 kW
Flow
0 m³/h
Pressure
0 bar
Tank Level
62.2%
Solar Irradiance
0 W/m²
Estimated PV Power
0.0 kW
Cloud Cover
98%
Temperature
21.2 °C
Energy Today
0.01 kWh
Water Today
0.11 m³
EC200U Link
Waiting
Modbus RTU
Waiting

Digital Twin Operating Mode

Select how the Digital Twin VFD should operate.

ACTIVE: VALIDATION
Live Architecture
Loading...
Active Digital Twin Source: Loading...
Digital Twin Fidelity
Physical INVT VFD versus Synthetic VFD
WAITING
Digital Twin Fidelity
Last synchronization
Physical asset Waveshare → INVT GD20
Digital Twin Synthetic VFD
Confidence is calculated by comparing the physical INVT VFD with the Synthetic VFD in Validation Mode.
Parameter Physical Twin Δ Status
Select Validation mode to view live comparison.
AI Model Validation
✓ Frequency behaviour closely matches the physical INVT GD20.
△ DC bus, motor voltage and motor current require Pilot-1 electrical calibration.
○ Power is currently a Digital Twin estimate.
✓ The Digital Twin does not write commands to the physical VFD.
Live Operator View

Solar Pump AI Control

Live operating condition and side-by-side control-strategy comparison.

LIVE
Pump RUNNING WAITING_FOR_SOLAR
Solar Irradiance 0 W/m² Night
Tank Level 62.25 % 12.45 m³
Applied Frequency 0 Hz Flow 0 m³/h

Control Strategy Comparison

All three values are calculated from the same current operating condition.

Active:
1

Baseline

Hz

Fixed reference frequency used as the comparison benchmark.

Reference control
2

Rule-AI

Hz

Engineering-rule recommendation based on available solar power.

Deterministic engineering logic
3

RF-AI

Hz

Pilot-1 Random Forest prediction based on current operating and weather inputs.

Simulated Benefits

Strategy Benefits Comparison

Affinity-law estimate
Strategy Frequency Estimated flow Estimated power Flow vs baseline Energy vs baseline
Baseline Reference -- 100.0% 100.0% Reference Reference
Rule-AI Engineering rule -- -- -- -- --
RF-AI Random Forest -- -- -- -- --

Estimated flow follows the frequency ratio. Estimated pump power follows the cube of the frequency ratio. These values are simulation estimates and are not measured field savings.

Predictive Analysis

Expected Performance by Sunset

LIVE PROJECTION
Tank Full ETA -- Current filling condition
Remaining Daylight -- Until local sunset
RF Water by Sunset -- m³ Projected cumulative water
RF Energy by Sunset -- kWh Projected cumulative energy
Strategy Projected water Projected energy Water vs baseline Energy vs baseline
Baseline Reference -- -- Reference Reference
Rule-AI Engineering rule -- -- -- --
RF-AI Active Random Forest -- -- -- --
Estimated Energy Saved -- kWh --%
Water Difference -- m³ --%
Estimated Cost Saved ₹ -- At ₹12 per kWh

Predictions assume the present operating condition continues until sunset. Tank auto-empty cycles may allow pumping to continue after the tank first reaches full level.

Operational Health

Pump-System Health Status

GOOD
Health Score --/100
Assessment Confidence --%
VFD fault status --
Motor current --
Motor voltage --
DC bus --
Frequency and flow stability --
Data and weather services --
RF model readiness --
Analysing current system signals…

This score represents operational health from available electrical, hydraulic, communication and model signals. Mechanical bearing and insulation condition require dedicated vibration and temperature sensors.

Frequency Application Status
AI-selected target -- Hz
Synthetic VFD applied -- Hz
Display source --
Physical VFD write Not enabled

RF-AI controls the synthetic Digital Twin only. No physical INVT GD20 write is performed from this dashboard.

Digital Twin Experiment

Test a Frequency and See the Impact

SYNTHETIC ONLY
RF trained-model recommendation Hz
Hz
RF-AI automatic control is active.
Predicted impact at Hz
Estimated flow -- m³/h
Estimated power -- kW
Flow vs RF-AI --%
Energy vs RF-AI --%
Strategy target -- Hz
Effective target -- Hz
Synthetic applied -- Hz
Physical VFD NOT AFFECTED
RF model basis and trained-model information
Model --
Status --
Raw prediction -- Hz
Safety-clamped prediction -- Hz
Model consensus --%
Safety range 20–50 Hz
Training period Loading…
Last retrained Loading…

Current model inputs

Loading…

Training performance

Training records --
Decision trees --
--
RMSE -- Hz
MAE -- Hz

Confidence shown here is agreement among the Random Forest trees. It is not a guarantee of field accuracy. The trained model file remains private on the edge controller.

Explainable AI

RF-AI recommends Hz

Confidence %
Measured irradiance 0 W/m²
Measured cloud cover 98 %
Measured tank level 62.25 %
Model status

The Random Forest recommendation is generated from current irradiance, DC bus voltage, solar score, temperature, cloud cover and wind inputs.

Current Decision
0.0 Hz

Against baseline Calculating…
Applied to Digital Twin 0 Hz

Comparison values are simulation outputs. They do not command the physical Pilot-1 VFD.

Advanced Controls and Fault Simulation

Engineering Controls

Power source: SOLAR_ONLY


Tank Validation Tools
Developer Interfaces

State API: /api/state  |  History API: /api/history

Live Trends

Frequency

PV Power

Flow

Tank Analytics

Synthetic tank model used for validation
Validation Mode
Tank Capacity
20.0 m³
20000 litres
Current Volume
12.45 m³
12450 litres
Remaining Capacity
7.55 m³
7550 litres
Current Cycle
#1
Started 2026-07-21T22:54:40
Cycle Elapsed
1h 8m
Estimated Time to Full
Waiting for flow
Last Fill Duration
Not completed yet
Last Full At
Not completed yet

Recent Fill Cycles

Cycle Mode Started Full At Duration Water Pumped
Waiting for the first completed fill cycle.

Tank Level

Day and Weather Details

Day status Night
Sunrise 06:02
Sunset 18:49
Humidity 87%
Wind 13.1 km/h
Weather API Connected

EC200U Connection Readiness

Virtual GD20 Ready
DTECH adapter Checking
Modbus RTU slave Checking
EC200U Checking
AI-Kit Cloud Not configured
Last request None

Virtual INVT GD20 Register Monitor

Export CSV
Loading register map...

Event and Alarm Log

Waiting for events...