SkyMes AI Vision

Automated Defect Detection with Computer Vision

The factory that sees, follows, and recognizes

It detects objects and anomalies, follows their movements, and recognizes the same item across different areas and cameras.

SkyMes turns images and video into production events

The three core capabilities

3) Recognize (Object Re-Identification — ReID)

Understand whether it is the same object.

SkyMes maintains an object’s identity when it disappears, reappears, or moves in front of a different camera.

Multi-camera systems can assign a global identity to objects observed by different cameras by combining visual features with spatial and time information.

Visual examples:

  • Same ID across cameras

  • Time and location aware

  • Robust to occlusions and viewpoint changes

1) See (Object detection)

Understand what is in the scene.

Parts, products, people, vehicles, containers, materials, defects and machine states.

Object detection identifies where an object is located and which category it belongs to.

Visual examples:

  • Boxes

  • Forklift

  • Missing component

  • Material detected

  • Person in restricted area

2) Follow (Object tracking)

Understand where it goes and how long it stays.

SkyMes follows the movement of objects, counts passages, and detects stops, queues, deviations and build-ups.

With object tracking, each object keeps its own ID while moving through the video.

Visual examples:

  • A line following a pallet

  • A timer above a container

  • A counter showing products passing through

  • An abnormal stop highlighted

  • A heatmap showing forklift routes

Use-case gallery

  • Quality

    Stop defects before they become waste.

    Discrete manufacturing: missing components, incorrect assembly, scratches, defective welds, non-compliant parts.

    Continuous manufacturing: color variations, visible contamination, uneven coating, faulty seals, surface anomalies.

  • Flow and productivity

    Count, follow, and measure the real production flow.

    Discrete manufacturing: work in progress between stations, queues in front of a machine, throughput time, stopped parts, or items moving along the wrong route.

    Continuous manufacturing: conveyor build-ups, empty lines, flow speed, counting cans, bags, rolls, or packages.

  • Visual traceability

    Maintain identity even when the code is not visible.

    Discrete manufacturing: follow the same frame, pallet, carrier, or container across different production cells.

    Continuous manufacturing: follow coils, reels, big bags, drums, tanks, or containers associated with a production batch.

  • Operational safety

    Detect a dangerous situation while it is happening.

    Discrete manufacturing: a person inside a robotic cell, a forklift close to an operator, access to a restricted area, missing personal protective equipment.

    Continuous manufacturing: presence in hot or controlled areas, obstacles along a passageway, vehicles in loading areas, visible spills.

  • Internal logistics

    Know what enters, what leaves, and where it goes.

    Discrete manufacturing: pallets at loading bays, containers being moved, forklifts, components delivered to the production line.

    Continuous manufacturing: trucks, tanks, drums, big bags, incoming raw materials, and outgoing finished products.

  • Process monitoring

    Turn a visible condition into a production event.

    Discrete manufacturing: machine running or stopped, empty feeder, tool present, jammed component, occupied workstation.

    Continuous manufacturing: material present on a conveyor, visible level, foam, overflow, flame, or an indicator in an abnormal position.

Technical details

From the camera to action

  • Images and video from the production line.

  • It detects, follows, and recognizes.

  • Defect, count, passage, stop, presence, or anomaly.

  • Alert, non-conformity, traceability update, task, or dashboard.

Integrated with MES, QCMS, WMS, CMMS, traceability and enterprise systems.

Show us what you need to see

Evaluate your use case

We start with a few images or a short video of your production line to evaluate the objects, movements, environmental conditions and process integration.

FAQ

  • Scratch, stain, crack, missing parts, foreign particles, surface deformation, discoloration, misalignment, custom setups are also possible

  • Yes, “good” baseline images and representative defective samples help train and tune the model. Our team assists with that.

  • Controlled lighting is important. We provide guidelines and calibration tools. In tricky cases, specialized lighting (e.g. coaxial, dark-field) or filters can help.

  • Yes, with adequate hardware, detection is near real-time. In other cases, batch mode is supported.

  • Yes, via APIs / SDKs / webhooks. The module works inside SkyMes’ integrated ecosystem.

  • User validation (accept/reject flagged defects) feeds back into retraining. Also periodic re-calibration and model versioning are supported.

  • Vision system include health monitoring, fallback inspections, and alerts when camera/lens issues occur.

  • In principle yes, but some products (very glossy, transparent, extremely fine textures) may present challenges. We’ll evaluate feasibility in your context.