Competences

The development of process automation and the consequent increase of productivity are currently associated with the better accuracy of quality control systems thus representing one of the most important prerogatives to acquire the minimum levels of competitiveness in the market. Machine Vision must be considered as one of the most efficient and used solutions, thanks to its applicability and integration with different production sectors.

The Loccioni Group in collaboration with robot producers has realized a series of solutions for quality testing in the production line using video cameras and anthropomorphic robots, able to check a high number of tests in a short cycle time.

close

Vibration testing of mechanical systems with laser vibrometers and/or more conventional accelerometers enables discrimination between good and faulty products and hence the analysis of vibration signals can be used for quality control on production lines. Normally, different manufacturing defects show different kinds of noise and vibration.

The nature of the problem is quite complex and requires the combination of advanced sensor technology and highly sophisticated data analysis techniques. Typical test stations employ advanced techniques of Pattern Recognition integrated into a complete system of data acquisition and signal processing. Complete and profound test on the final products is an essential requirement to obtain the necessary information for an adequate control of the production process.

close

The field of robotics comprises a broad spectrum of different technologies in which computational intelligence is embedded in real machines, creating systems with capabilities far exceeding the core components alone. Such robotic systems are then able to carry out tasks that are unachievable by conventional machines, or even by humans working with conventional tools. The ability of a machine to move by itself is one such capability that opens up an enormous range of applications uniquely suited to robotics.

The use of robot manipulators to make flexible non-contact sensors or cameras and to place them at specific locations has the purpose to vary sensors position during test phase in order to adapt to changes in environment lay-out or position and geometry. In fact, most non-contact sensors need to be placed at specific positions in space relative to the measured object. Deploying of robots to handle sensors allows their re-positioning, so to keep signal-to-noise at better levels.

close

Most current applications see a robot operating with a human user or with a collection of sensors in a very structured environment in a previously determined manner. With the emergence of networked, embedded systems and the increased presence of networks in homes and in factories, robots can work with other robots, learn from different types of sensors and interact with different human users. This is particularly true for mobile robotic systems whose environments are constantly changing.

The use of mobile platforms to build semi or fully autonomic diagnostic stations has the purpose to avoid the replication of expensive testing equipment. The robot can move even in a low structured environment and reconfigure the measuring task. The availability of a mobile diagnostic platform also pursues the objective of ensuring a high standard for repeatability and reproducibility of the tests.

close

Automatic Classification (AC) deals with the application of advanced signal processing methodologies for Quality Control (QC) in production, whose aim is to determine the conformity of a product to design specifications or regulations. Therefore starting from measurement data AC assigns a class to the product which defines its overall “quality”.
AC is here intended as the whole set of software and signal processing algorithms that, having as an input the measured signals, determines autonomously – i.e. without human intervention – the quality class the object belongs to.
The editing and updating of test plans is actually a long and difficult task requiring a deep knowledge of the product under test as well as of the measurement system. For this reason one of the main efforts of the R4I team is on the development and implementation of innovative self learning algorithms for automatic update of test plans, in order to face with new or different models of a product, avoiding system drifts.

close

Machine Vision Framework is a software platform developed by the Loccioni Team for the acquisition and processing of images taken from a single or multiple cameras. In order to develop a complete system for the analysis of different defects, several imaging processing tools have been implemented, such as: Find Edge algorithm which helps to find the object contours on the image; Match Pattern algorithm, with which it is possible to find a template on the image through cross-correlation; Threshold algorithm which transforms the image in a binary one; Caliper algorithm which helps to measure the distance among points on the image; Read Digits/Digital Code algorithm which allows to check the status of the bright elements (e.g. LEDs, displays) of many instruments interfaces; Read Barcode/Text algorithm, with which it is possible to read text or products barcode on the image. Moreover there are classical image processing algorithms, such as equalization, segmentation and filtering.
This kind of flexibility makes this software very suitable for robotics, since it allows at the same time to “teach” some tasks to the robot and to execute them in the operating environment.

close

Fundamental prerequisite for industrial application of measurement device is the availability of dedicated calibration methods suitable for evaluating and adjusting the precision and accuracy of measurement equipment. Calibration procedure allows the estimating of the sensor uncertainty in the specific test condition which is similar to in-line application. All testing systems developed by Loccioni are designed for capability to be calibrated for a specific need and application, which includes the equipment settings, the transducer, and the test setup, in order to validate that the desired levels of precision and accuracy are achieved.

The proposed strategy for calibration is based on the ISO15530-3 Guide of Expression of Uncertainty in measurement (GUM) and it developed a calibration procedure using ‘Type A’ approach. It contains a process for determining the calibration of the sensors by statistical treatment of experimental data by measuring calibration standard traceable back and in accordance with the National Institute of Standards and Technology (NIST).

close

Multivariate analysis (MVA) denotes a set of techniques dedicated to the analysis of data sets which involves observations of more than one variable. In design and analysis, the techniques are used to perform trade studies across multiple dimensions while taking into account the effects of all variables on the responses of interest.

Multivariate Statistical Process Control (MSPC) refers to a set of monitoring and control advanced techniques of the operating performance of a batch or a continuous process, when several related process variables are of interest.

close