MV-VIEW-01 CatalogueVIEW-02 Station commissioningVIEW-03 SDK & driver compatibilityVIEW-04 Capture & evidenceVIEW-05 Trigger & capture groupsVIEW-06 CalibrationVIEW-07 Source contract & mappingVIEW-08 Inspection & provenanceVIEW-09 Vision / WMS / LocateLinkVIEW-10 Fault & recovery labVIEW-11 Qualification & coverageVIEW-12 Release & rollbackEyeZ job catalogue (60)Software & standards (16)Primary sources (54)Import checks
Acting asAuthorization is evaluated by effect, not by screen. Demonstration only.0 session operations
EYZ-MV-037
Random-bin pick-pose proposal
ProposedRobot guidance and pick verification← EYZ-MV-036EYZ-MV-038 →Synthetic vision workspace. No camera, frame grabber or vision controller has been contacted; camera SDKs run only in engineering-managed edge workers, never in a browser. No real image exists here — every frame is a described placeholder. Nothing is live or qualified, and no inspection result releases product or causes any robot, reject, dispatch or stock effect.
The job
Actor
Robotics/integration engineer
Job to be done
Identify a candidate part and pose for a separately governed robot planner.
Capture requirement
Qualified 3D cloud, camera-to-robot transform and occlusion handling.
Required context
Robot cell, bin, object, gripper, robot/world coordinate frames and authority boundary
Responsibility and proof
EyeZ output
Candidate pose, object ID and confidence/uncertainty.
CoreSync responsibility
Acquire the native observations and evidence references; preserve source identity, capture time/epoch, quality, recipe/calibration versions and delivery evidence. Bind through approved canonical semantics; do not infer control authority.
Acceptance evidence
Evaluate pose error and successful perception under clutter; robot tests separate.
Guardrail
Perception output is not a robot command or proof of collision-free motion.
Candidate camera profiles (7)
| MV-CAM-008 | LUCID Vision Labs Arena-supported GigE cameras | Arena SDK | DETAIL |
| MV-CAM-015 | Ensenso / IDS Ensenso stereo camera families | Ensenso SDK NxLib | DETAIL |
| MV-CAM-016 | Zivid Zivid 3D cameras | Zivid SDK | DETAIL |
| MV-CAM-017 | Photoneo PhoXi and MotionCam-3D | PhoXi Control API | DETAIL |
| MV-CAM-018 | LMI Technologies Gocator Classic / GoPxL | Gocator SDK / GoPxL interface by device generation | SURFACE |
| MV-CAM-022 | Stereolabs ZED camera families | ZED SDK | DETAIL |
| MV-CAM-032 | ifm O3R vision platform | ifm3d | DETAIL |
A candidate is a design reference from a reviewed technical source. It is not an implemented capability, not coverage and not a qualified adapter.
Candidate software and standards (4)
- MV-SW-005 MVTec HALCON / MERLIC interfaces — Vision-processing and acquisition integration; gate: Tool licensing and exact acquisition-interface revision required; not every camera is qualified.
- MV-SW-009 OpenCV — Image-processing library; gate: Audit package and third-party dependencies; OpenCV 4.5+ license differs from older releases.
- MV-SW-010 OpenVINO — Inference runtime; gate: CPU model/operator/precision coverage and latency must be benchmarked; accelerators optional.
- MV-SW-011 ONNX Runtime — Inference runtime; gate: Operator set, preprocessing, numerical equivalence and provider fallback are qualification items.
Synthetic results attributed to this job (0)
No synthetic assertion is attributed to this job yet. Run a scenario in the fault and recovery laboratory.