- Works where all the others fail
PENSO® is designed to work on unpredictable variations and defects, as well as on objects difficult to model or without a golden reference.
- Fast configuration, even faster prototyping
PENSO® full configuration requires hours, not weeks and produce feasibility studies in minutes, not days, without any programming required.
- Adapts easily to the real working conditions
PENSO® helps the consistency of the inspection process by autoadjusting to changes on working conditions, self-learning from the ongoing process.
PENSO® is an artificial intelligence-based computational unit for imaging applications. PENSO® self-learns the expected features of an object by simply looking at a small series of samples, regardless of the possible presence of defective product in the midst. Once trained, PENSO® will start operating autonomously and warning you if something unwanted happens.
Object modeling, often very difficult or even impossible is therefore no longer needed; moreover, you are not asked to define possible object variations or defects from the beginning, usually impossible. Instead, as a human operator would do after observing some samples, PENSO® understands what will be considered normal and acceptable from now on. Since neither modeling or programming is needed, with just some elementary configuration tools you can immediately understand if an application is feasible or not and really save a lot of time. Then, with minimum additional effort, you can fine tune the application and make it fully working.
|Description||Artificial Intelligence-based Vision Unit|
|Camera (1)||Matrix Vision up to 5 MPx , 3D cameras (see Accessories)|
|Number of parts per second (2)||20|
|LED indicators||Power, Status and Error|
|Number of storable images (3)||≈ 800K|
|Synchronization inputs n°||2, opto isolated, common reference|
|Commands||6, opto isolated, common reference|
|Status||4, opto isolated, common reference|
|Synchronization outputs n°||2 strobe trigger, opto isolated, common reference, 1 camera trigger, opto isolated|
|Results||6, opto isolated, common reference|
|Elaborated||1, opto isolated|
|USB 3.0||2 (camera dedicated)|
|Ethernet||1 (camera dedicated) + 1|
|1 PS/2||1 (keyboard and mouse)|
|Voltage||V, DC||24 ± 5%|
|Max power consumption||W||100|
|Humidity||20-85% (with no condenstation)|
|IP rating||no IP class|
|Installation||indoor use only|
- Non-standard cameras available upon request (customization).
- Estimated value. The number of inspected parts per second may vary depending on image resolution, line speed and number of activated features.
- Estimated value based on 250 Kbytes images stored in 200 GB SSD memory.
PENSO® is extremely easy to install: you simply need to connect inputs (camera, keyboard/mouse etc.) and outputs (synchronization signal, OK/NOK signal, etc.) and provide 24V DC.
The basic settings are extremely simple and fast: image brightness, color correction, focusing and segmentation (i.e. is the process of separating the product to be inspected from the background) are assisted by convenient software tools. The interaction with PENSO® is possible at all time through standard physical interfaces (keyboard, mouse and screen) or remotely, integrating PENSO® within the local network.
The learning process is easily performed by presenting some products on the production line and activating PENSO® in “LEARN” mode during normal operation.
Unlike traditional vision units, PENSO® autonomously learns the characteristics of your production in a few minutes: it is normally sufficient to present a few tens or a few hundreds of products during production to allow PENSO® to learn their characteristics without complicated settings.
PENSO® can tolerate up to 20% defective products during the learning phase, without affecting its ability to sort products correctly. PENSO® will be ready to check your production once the status bar is full.
Moreover, whenever the goods on your production line change or anytime you want to adjust your quality control process to new production parameters, you can just press the “LEARN” button and PENSO® will adjust itself accordingly.
Even during the learning phase, PENSO® continues to monitor production, quickly adapting to the new inspection criteria without having to stop the line: no other vision system is so flexible and easy to configure.
Once the learning process is complete, PENSO® is ready for the sorting phase or “CHECK” phase: the products deemed inconsistent with the desired level of quality are reported and can be rejected from the line by interfacing PENSO® with most common ejection systems thanks to the preinstalled opto-isolated outputs.
PENSO® can store images of defective products, also keeping track of the reasons for rejection: this data can then be analyzed to improve the production process.
You can also adjust the “severity” level of the control parameters without having to stop the line. A dedicated slider bar allows the user to loosen or tighten the sorting criteria, quickly and easily adjusting PENSO® to new quality parameters.
PENSO® also allows a higher level of control: we can decide which features describing the object color and shape are taken into consideration by PENSO®, and how much PENSO® will consider them critical for the final decision (weight), all done in real time.
PENSO® is successfully applied in many imaging applications.
Here are some of the many cases where AI-driven imaging is the key to success.
Color/Mono imaging: Dishwasher tablets
Dishwasher tablets come with natural variations: the density of blue over white soap, the roundness of the red core, etc. Nevertheless, PENSO® sees the defects as deviation from the average product shape or color distribution, exactly the way a human operator would.
Color/Mono imaging: Tealight candles
Inspecting candles is not an easy task: surface features and texture are always slightly different, making it hard to work within the usual deterministic imaging approach. An AI-driven inspection, on the other hand, skips the unimportant details and shows us only what matters!
The typical output of a 3D camera is a monochrome picture of the object, where dark regions are lower and bright regions are higher (height map). In this example, a package with broken or irregular crackers is first detected by PENSO® because of its unusual dark spots, corresponding to the missing chunks. Too many crackers in the package, instead, appear as a uniform brighter-than-average tint, corresponding to the increased height.
When setting up the image for PENSO®, we want to make sure it represents the product as accurately as possible. Therefore, our first step is system calibration: an intuitive tool guides the user through the optimization of focusing, color calibration and light intensity
In our second step, we need to tell PENSO® how to separate the object from the surface underneath (segmentation). This goal is also achieved using a simple software tool, which lets us play with intensity thresholding and color differences to quickly and easily get the best result.
Feature extraction management
Once we set our best image and the LEARN process is completed, we can start fine-tuning the sorting process, selecting from a growing number of features the ones that best relate to our specific sorting process.
PENSO® will show us in real time how a single feature and its relative importance (weight) contributes to the overall choice between OK and NOK sample.
- Do you need a non-standard feature? We will create a it for you!
- Contact us to know about customized filters in PENSO®