SYNOPTARA

SYNOPTARA

Industrial Vision Intelligence

Italiano · Deutsch · Français

SYNOPTARA is a standard software platform for industrial visual inspection, combining multi-camera acquisition, deterministic machine vision, dimensional measurement, and artificial intelligence in a single operating environment.

The platform is built around a clear principle: provide a broad set of reusable inspection capabilities without rebuilding the software for every machine, component, or production process.

Cameras, recipes, localization strategies, geometries, measurements, tolerances, AI models, results, diagnostics, and machine communication share the same common structure.

One platform, multiple inspection tasks

SYNOPTARA supports up to 16 cameras, operating independently or organized into functional groups.

Each camera can manage its own:

• acquisition parameters;
• localization strategy;
• inspection tools;
• dimensional measurements;
• surface controls;
• AI models;
• tolerances;
• diagnostic information;
• inspection results.

Results from different cameras and inspection stages can be combined into a coordinated decision for the inspected part.

This allows the same software platform to cover applications ranging from a single inspection station to complex systems observing the product from multiple viewpoints.

AI-guided metrology

SYNOPTARA turns AI detections into traceable dimensional measurements by combining neural inference with deterministic machine-vision metrology.

Segmentation models locate components, regions, and complex features even when their position, orientation, appearance, or background changes.

The resulting masks can be transformed into geometrical features such as:

• edges;
• contours;
• lines;
• arcs;
• circles;
• centers;
• reference points.

These AI-detected features are then refined through mathematical and conventional machine-vision algorithms and used to construct dimensional measurements.

AI determines where and what to measure. Deterministic geometry calculates the final value.

This hybrid approach combines the adaptability of artificial intelligence with the precision, repeatability, and explainability required in industrial inspection.

Measurement stability still depends on controlled image acquisition, lighting, calibration, mechanical repeatability, and the robustness of the trained model.

Standard does not mean limited

SYNOPTARA is not intended to be rewritten for every application.

Inspection capabilities are selected and combined within a common structure. Cameras, localization tools, geometrical features, measurements, tolerances, surface controls, and AI modules remain part of the same operating model.

This approach reduces application-specific code, simplifies maintenance, improves consistency between machines, and makes inspection recipes easier to verify, reuse, and evolve.

What SYNOPTARA simplifies

• multi-camera inspection management;
• visual recipe configuration;
• part localization before measurement;
• construction of geometrical references;
• dimensional measurement from traditional and AI-detected features;
• surface inspection and anomaly detection;
• acquisition of images for datasets and model development;
• testing and offline simulation;
• diagnostics and result traceability;
• controlled access to technical functions;
• communication with automated machinery;
• coordinated inspection decisions.

Industrial workflow

A typical SYNOPTARA inspection follows a structured process:

1. image acquisition;
2. part localization;
3. extraction of relevant features;
4. construction of geometrical references;
5. execution of measurements and visual controls;
6. comparison with tolerances;
7. coordination of results from multiple cameras;
8. final decision for the inspected part;
9. communication of results to the machine.

The same structure is used during setup, testing, and production.

Designed for industrial use

SYNOPTARA is intended for machine builders, system integrators, and manufacturers working with automated inspection and high-throughput production systems.

The platform is based on hands-on experience with industrial vision systems dating back to 1999.

Its development focuses on:

• operational clarity;
• repeatable configuration;
• controlled access to sensitive functions;
• local data management;
• CPU and GPU processing;
• fault diagnostics;
• maintainability;
• reuse of validated inspection methods.

Current capabilities

Capability	Current status
Configurable cameras	up to 16
Functional camera groups	up to 10
Localization strategies	8 strategies in 5 families
Dedicated standard shapes	2, 4, 6, and 8 faces
Processing profiles	CPU and GPU
Inspection methods	geometry, surfaces, classification, segmentation, anomaly detection
Standard-parts catalog	local, editable, traceable, and versionable
Operating modes	configuration, simulation, testing, and production

Current project status

Version 1.01 includes the main platform functions and is available for technical validation using real images, representative samples, and industrial hardware.

Machine testing is used to verify:

• image quality and stability;
• localization robustness;
• measurement repeatability;
• defect coverage;
• false-positive and false-negative rates;
• processing time;
• CPU and GPU load;
• behavior during camera or image faults;
• recipe changes and restart recovery.

The platform provides a standard technical foundation. Final inspection performance must always be validated under the actual production conditions.

Vision

To provide a common software core for industrial quality-control systems capable of observing a product from multiple viewpoints, applying different inspection methods, and transforming industrial images into reliable decisions that automated machinery can use.

Preliminary validation

Download the six-camera test brochure (Italian, PDF)

Download the multi-camera AI test brochure (Italian, PDF)

Product documentation

Explore SYNOPTARA capabilities and technical documentation

The public documentation covers:

• multi-camera operation;
• functional camera groups;
• localization strategies;
• geometrical features and measurements;
• AI-assisted inspection workflows;
• standard-parts catalog management;
• architecture and machine integration;
• security and operational roles;
• technical validation.

Proprietary source code, internal implementation details, private datasets, and reserved technical information are not published.

Contact

For information, technical discussions, or project inquiries:

develop@synoptara.dev

‍SYNOPTARA is under private development. Source code and proprietary implementation details are not published in this repository.