— 05 · Logistics

Perception, mission and operational intelligence

Logistics operations depend on knowing where assets are, in what state, and what the operation should do next. Convex develops the visual perception, localization, mission planning and telemetry layers that support that tracking.

Convex does not sell a logistics automation catalogue. What we build are the layers any physical operation depends on: visual perception, localization, mission planning, telemetry and operational intelligence about what is actually happening.

Perception, mission and operational intelligence
05 · Logistics
— Operational context

The operational problem

  1. 01
    Asset state out of date

    Where an asset is and in what condition is usually inferred rather than observed.

  2. 02
    Limited environment perception

    Physical operations change layout and occupancy continuously, and the system has to see that.

  3. 03
    Missions with no queryable state

    Without a mission layer, each task becomes an isolated event with no trail.

  4. 04
    Scattered telemetry

    Position and condition data live in separate systems and never form a single view.

  5. 05
    Decision detached from execution

    What the software decides and what happens on the floor rarely share the same record.

— Convex architecture

From observation to action

The same architecture runs across every sector. What changes is which layer is critical and which proprietary systems take part.

Observe
Interpret
Coordinate
Act
Observe

Nexa Vision

Visual perception of the environment, identification and tracking of assets.

Interpret

Nexa Vision

Reading of what was observed: presence, relative position and change of state.

Coordinate

Synapse Control

Mission planning, routing, telemetry and aggregate operation state.

Act

Axis

Physical execution of the planned task, returning state to coordination.

— Data layers

What feeds the pipeline

Sources compatible with the systems deployed in this sector. No layer is assumed: every monitored area declares what is actually available.

Cameras

Visual streams over points of interest in the operation.

Images

Captures bound to identification and verification events.

Localization data

Relative position of assets within the observed environment.

Telemetry

State, condition and movement of the platforms.

Mission state

Planned route, current step and execution result.

Operational history

Queryable record of past events and missions.

— Applications

Technical applications

Each application states the problem addressed, the data used, the Convex system involved and the operational output produced.

Visual asset tracking

Location and state inferred instead of observed.

Data
Camera streams and images per event.
System
Nexa Vision
Output
Record of presence, relative position and change of asset state.

Navigation and localization in dynamic space

Layout and occupancy change through the shift.

Data
Visual perception and localization data.
System
Nexa Vision + Synapse Control
Output
An up-to-date spatial reference for mission planning.

Mission planning and coordination

Tasks executed without queryable state.

Data
Planned route, telemetry, platform state.
System
Synapse Control
Output
Mission queue with aggregate state and a result per execution.

Physical execution tied to the decision

Distance between what the software decides and what happens on the floor.

Data
Mission state and environment perception.
System
Axis
Output
Action executed with state returned to the same record.

Operational intelligence over history

Isolated events with no aggregate reading.

Data
Mission history, telemetry and visual events.
System
Nexa M1
Output
Consolidated reading of what happened over the observed period.
— Operational flow

One cycle, from sensor to record

  1. 01Capture

    Visual perception records the environment and assets at the point of interest.

  2. 02Processing

    Identification and relative position are extracted from the observation.

  3. 03Interpretation

    Observed state is compared against the state the mission expects.

  4. 04Decision

    Divergences raise an alert and reorder the mission queue.

  5. 05Action

    Synapse Control coordinates the step; where applicable, Axis executes physically.

  6. 06Record

    Mission, event and result are archived for later query.

— Deliverables

What the system delivers

Alerts

Divergence between observed and expected state.

Telemetry

Position and condition of the platforms in operation.

Planning

Mission queue with routing and sequencing.

Operational history

Queryable record of missions and events.

Reports

Period consolidation of what was executed.

— Technology stage

Maturity by capability

We separate what is in operation from what is still under development. Research and prototypes are never presented as deployed product.

  • In operation
    Visual perception and asset identification

    Observation layer applied to operational environments.

  • In development
    Mission planning and telemetry

    Route coordination and aggregate operation state.

  • In development
    Autonomous navigation in dynamic space

    Localization and continuous replanning under research.

  • In development
    Physical execution with Axis

    Robotic actuation integrated with the mission layer.

— Application contexts

Application contexts

Distribution centres

Dense environments with constantly changing layout and occupancy. An application context, not an off-the-shelf product.

Yards and outdoor areas

Visual tracking of assets across wide, lightly instrumented spaces.

Internal movement

Repetitive trips between fixed points, well suited to planned missions.

Verification and dispatch

Checkpoints that need a traceable visual record.