VARMAXAquaTech Platform
Execution

Prove the core field-control loop, then add intelligence and autonomy

The most efficient path is to prove reliable gateway-to-node control before adding sophisticated sensing or autonomous flight. Future modules must attach without redesigning the platform.

13 · Development Roadmap

Phased execution plan

Each phase carries its own exit criterion.

  1. 0 — Architecture
    Exit: All modules have clear contracts

    Workspace, protocols, diagrams, interfaces, safety boundaries

    CURRENT PROTOTYPE
  2. 1 — Device Prototype
    Exit: Stable bidirectional communication

    Primary gateway + one secondary + LoRa messaging

    CURRENT PROTOTYPE
  3. 2 — Motor Control
    Exit: Verified ON/OFF and failure behavior

    Safe integration with representative control/load

    CURRENT PROTOTYPE
  4. 3 — Backend
    Exit: End-to-end API workflow

    Auth, farms, ponds, devices, commands, telemetry/events

    PLANNED
  5. 4 — Mobile App
    Exit: Core farmer workflow works

    Pairing, farm setup, pond view, control, schedules, alerts

    PLANNED
  6. 5 — Web
    Exit: Operational dashboard usable

    Admin/support/device management and analytics

    PLANNED
  7. 6 — AI Orchestration
    Exit: Repeatable multi-step automation

    Planner, task graph, context, concurrency, audit, safety gate

    PLANNED
  8. 7 — Pilot
    Exit: Pilot metrics achieved

    Real farm deployment and fault/UX analysis

    PLANNED
  9. 8 — Future Sensors
    Exit: Validated measurements

    DO and electrical telemetry prototypes

    FUTURE
  10. 9 — Drone
    Exit: Controlled field mission validated

    Feed hardware + mission software + safety validation

    FUTURE
14 · Future DO SensorFUTURE

Dissolved oxygen sensing — future track

Feasible as a future subsystem, but accuracy, calibration, temperature compensation, fouling, maintenance and field validation are the dominant engineering challenges.

  • Electrochemical or optical DO probe
  • Analog front-end or digital probe interface
  • Temperature measurement for compensation
  • MCU + LoRa communications
  • Calibration workflow and history
  • Local filtering and plausibility checks
  • Battery/power management for pond deployment
Hard constraint

Do not drive automated aerator decisions from raw probe voltage without calibration and field validation.

15 · Future Current / Voltage MonitoringFUTURE

Electrical telemetry — future track

The project explicitly decided not to include current and voltage monitoring in the current implementation messaging. The architecture can reserve interfaces for it later.

  • Motor runtime verification and fault/overload analytics
  • Energy consumption and maintenance insight
  • Appropriately rated isolated sensing hardware and protection
  • Electrical measurement separated from the low-voltage MCU domain
  • Backend distinguishes commanded state from measured electrical state

ON command ≠ proof that the motor is actually running.

16 · Autonomous Drone Feed DistributionFUTURE

Feed missions over multiple ponds — future concept

Treated as a separate safety-critical product track. It should not be coupled directly to motor-control logic or unrestricted AI actions.

Farmer selects ponds / feed plan→Backend validates schedule→Mission planner creates route→Drone distributes pond-by-pond→Telemetry / status returns→Mission verified→Feed log stored
Required subsystems
  • Drone flight controller/autopilot integration
  • Feed hopper and calibrated dispensing mechanism
  • Quantity estimation or dispensing feedback
  • GPS/positioning and geofencing
  • Obstacle and fail-safe strategy
  • Battery and return-to-home logic
  • Mobile mission control interface
  • Backend mission records and audit trail
Major risks & engineering decisions
  • LoRa collisions at scale
    Lost/delayed messages → Controlled transmission, bounded retries and radio testing
  • Electrical integration
    Equipment damage/safety risk → Qualified electrical design; retain industrial protection/isolation
  • Internet dependency
    Loss of remote access → Local gateway autonomy and offline queues
  • AI incorrect action
    Operational disruption → Typed commands + authorization + safety gate + verification
  • Drone operation
    High safety/regulatory complexity → Separate subsystem, geofencing, fail-safe and controlled pilots
17 · Closing

Strategic principle

First make the system reliable. Then make it intelligent. Then make it autonomous.