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
Monitor operating conditions, understand AI decisions, compare control strategies and launch specialised solar applications from a single client-facing platform.
Each application addresses a distinct operational need while remaining part of the Altius AI-Kit ecosystem.
Real-time solar-pump simulation, VFD telemetry, machine-learning recommendations and operational validation.
Analyse module images and operating information to prioritise cleaning and understand potential soiling-related losses.
Explore irradiance, weather and location scenarios to understand expected solar-system behaviour.
Centralised telemetry, pilot monitoring, analytics, documents and multi-site operational visibility.
The EC200U changes to Connected only after a real Modbus request reaches the isolated DTECH adapter.
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.
Select how the Digital Twin VFD should operate.
| Parameter | Physical | Twin | Δ | Status |
|---|---|---|---|---|
| Select Validation mode to view live comparison. | ||||
Live operating condition and side-by-side control-strategy comparison.
All three values are calculated from the same current operating condition.
Fixed reference frequency used as the comparison benchmark.
Engineering-rule recommendation based on available solar power.
Pilot-1 Random Forest prediction based on current operating and weather inputs.
| 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.
| Strategy | Projected water | Projected energy | Water vs baseline | Energy vs baseline |
|---|---|---|---|---|
| Baseline Reference | -- | -- | Reference | Reference |
| Rule-AI Engineering rule | -- | -- | -- | -- |
| RF-AI Active Random Forest | -- | -- | -- | -- |
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.
This score represents operational health from available electrical, hydraulic, communication and model signals. Mechanical bearing and insulation condition require dedicated vibration and temperature sensors.
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.
The Random Forest recommendation is generated from current irradiance, DC bus voltage, solar score, temperature, cloud cover and wind inputs.
Comparison values are simulation outputs. They do not command the physical Pilot-1 VFD.
Power source: SOLAR_ONLY
State API:
/api/state
|
History API:
/api/history
| Cycle | Mode | Started | Full At | Duration | Water Pumped |
|---|---|---|---|---|---|
| Waiting for the first completed fill cycle. | |||||