The problem/need

The need

There is a ubiquitous need for defect inspection and visualization in all manufacturing processes. Starting from industry, each factory has a dedicated quality control department that inspects manufactured goods, intermediate products or production equipment in order to ensure quality and production efficiency.

The problem

While automation for such tasks has certainly gained ground in recent years, quality control procedures still require human intervention to a large extent either for manual inspection or for verification of computer generated indications.

Our solution

Development and integration of IoT sensor networks

We provide customizable solution that utilizes 2D industrial cameras and 3D sensors for quality inspection. AUTOMAVIS will move beyond integration of commercial sensors, as is common in competition, and develop AI-enhanced IoT networks that can adapt acquisition parameters in order to optimize data acquisition. To this end, smart and adaptive illumination will be part of the installed IoT networks.

AI software for sensory data processing

We focus on Deep Learning methods for analysis of 2D and 3D inspection data. A modular software library will be developed incorporating deep computer vision methodologies for instance segmentation and detection that will optimized for the quality control tasks. The AI library will support deployments on both edge and cloud side in order to better adapt to different production lines according to customers’ requirements.

Ergonomic visualization of quality control results

Our solution provides results through user-friendly graphical interfaces that will notify human operators on identified defect while allowing them to rapidly explore inspection data. Through AUTOMAVIS visualization software, smart data annotation will be supported in order to accelerate the creation of the necessary training data for each customer. Depending on the customers’ requirements both desktop and web-based visualization will be supported with the appropriate interfaces with the IoT network and the customer IT backbone.

Application

As a company offering machine vision solutions for quality inspection, AUTOMAVIS has two categories of customers. The first one refers to industries and companies that have specific needs for quality control using machine vision while the second category is system integrators that play an intermediate role by incorporating AUTOMAVIS modules in their systems and selling them to interested entities.

MANUFACTURING INDUSTRIES

Industries that want to automatet their quality inspection process using machine vision

Car rental companies

Vehicle sites of car rental companies where vehicles are returned and automatic recognition of damages is needed

Logistics and good delivery companies

Vehicle fleet monitoring where the customer wants to monitor the status of its vehicles, in order to detect new damages and decide whether a vehicle can be put in circulation or has to be repaired

Vehicle repair workshop and dealerships

The goal is to identify defects on the surface of vehicles that could generate additional revenue for the customer

System integrators

Intermediate companies incorporate our solution for machine vision inspection and visualization in their own products, before selling