Part image acquisition
Every cycle starts with stable, repeatable acquisition. Magpie images the real part directly in the cell, using dedicated optics and lighting to make the most critical machined surfaces readable.
Die-cast part surface inspection after machining.
Explore the system →Magpie is the AI vision system for die-cast part surface inspection after machining. It inspects every part with centesimal precision, detecting defects invisible to the human eye, porosity, burrs, cracks, and other non-conformities directly in the work cell.
Magpie does not simply take a picture of a component. It reconstructs the machining context, isolates the truly critical surfaces, applies hybrid vision algorithms, and returns a traceable result ready for the line.
Every cycle starts with stable, repeatable acquisition. Magpie images the real part directly in the cell, using dedicated optics and lighting to make the most critical machined surfaces readable.
AI automatically identifies where the part has been machined and where non-conformities should actually be searched for. No generic inspection: analysis is focused only on the surfaces that matter.
On the detected surfaces, Magpie combines traditional machine-vision algorithms for precise rules and measurements with AI algorithms to classify complex, variable defects that are difficult to define in advance.
The system immediately returns the pass/fail result and stores images, applied criteria, and inspection history. Every decision remains traceable and available for reports, databases, or integration with factory systems.
Image. Understanding. Analysis. Result.
Click on the points on the part to discover each type of defect detected by Magpie.
↑ Click on the cyan dots to explore each defect
Select a point on the part
12 defect categories, one system.
Pure AI and traditional vision combined. The best of both worlds.
Advanced hybrid algorithms combining traditional rule-based vision for precise geometric measurements and Deep Learning for complex defect classification. The integrated NPU runs AI inference directly on-device.
Native connectivity with PLCs, SCADA and MES. Every part receives a digital quality certificate. Track & Trace per serial number with complete inspection history.
Move the vertical line to compare the raw image with the software analysis. Magpie highlights size, minimum distance, and aggregation to apply real acceptance rules.
Move the cursor to see how the software turns a raw image into a reject or accept decision rule.
Not a simple pass/fail sorter. A complete production controller, with data traceability built into every part that passes through.
Full production traceability, part by part. Real-time statistics, pass/fail counters, defect history by shift. Everything one click away, with no specialist training required.
Interface designed for foundry operators, not software engineers. Multi-station view, defect heatmaps, detection points mapped onto the part. Zero hidden complexity.
Acceptance thresholds, regions of interest, multi-parameter criteria: everything is configured directly from the interface. Switch tolerances, sensitivity and formats without touching a line of code.
Two Magpie system demo videos with an integrated overlay player, plus one application render viewable full-screen.
Inspection of machined surfaces on oil pan covers to ensure perfect sealing. Critical control of engine mating flanges.
Millimetric precision control on active safety components. Surface quality of machined areas is critical for vehicle performance and safety.
Magpie does not work alone. It is part of an integrated suite that monitors every critical die-casting phase, creating an uninterrupted quality chain.
Inspection of the casting just extracted from the press. Pin breakage, thermal regulation anomalies, lubrication issues, process stability. Analysis in under one second directly in the cell.
360° 3D inspection after trimming. Full dimensional verification, hole occlusions, flatness, integrity of the trimmed profile.
Final inspection of surface porosity, machining burrs and surface finish quality before shipment.
Everything you need to know before evaluating Magpie for your line.
Magpie is optimised for machined aluminium die-cast parts, but adapts to any geometry. The system is configured case by case through learning on master samples of the customer's specific part.
Magpie natively supports OPC UA, Ethernet/IP and Profinet. It integrates with any PLC or MES/SCADA system supporting these standards, including Siemens, Rockwell, Beckhoff and others.
Yes. Every criterion (size, numerosity, distance, concentration) is configurable through a software interface. Thresholds can vary by functional area, enabling custom multi-parameter rules for each application.
Yes. All inspected parts and related data are tracked in a database and can be made available externally or exported through reports, depending on the customer's needs.
Analysis and AI inference time depend on part geometry and on the number of checks required. In many applications processing can be completed in under one second, while more complex inspections may require different timings depending on the project. The integrated NPU runs AI inference directly on-device, ensuring real-time performance without dependency on external servers.
Timing depends on project complexity and integration with the existing line. Imago handles the entire installation, configuration, and commissioning process, supporting the customer through full production start-up.
The system is designed to operate in industrial environments and to keep routine maintenance to a minimum. Positioning is defined with Imago technicians to ensure long-term reliability. We recommend periodic checks of the panel condition and cleaning of optical surfaces whenever needed.
Imago takes great care of customer support. A training course is included during installation and can also be repeated later. We have a Service department available during Italian business hours and dedicated assistance packages for different needs.
Imago Vision is an Italian company specialized in advanced industrial vision systems. For more than 20 years it has developed automated quality-control solutions, combining traditional machine vision, artificial intelligence, and hardware designed for high-intensity production environments.
Magpie comes from that experience applied directly in foundries and machining lines. It is not a lab concept, but a system built to identify exposed porosity, burrs, surface defects, and dimensional anomalies at the most critical point: before the part is shipped.
See how Magpie can transform post-machining quality control in your plant. Each installation is tuned to the part, line throughput, and the real acceptance criteria used in production.
Or email info@imagovision.it and we will reply within 24 business hours.