*Quality of the inspection results:

A minimum of 8Mp photography  is mandatory. The quality of the photography is highly depended on the camera sensor / optics. If you are planning to test the Agnospcb automated optical inspection tool with a smartphone’s camera, we strongly recommend using high-end models such as iPhone 9/10/+, Huawei P20/30 or Google Pixel 3/4.

For the best possible detection results, we have designed an inspection platform based on the powerful Sony Alpha camera. You can replicate if by yourself or get it from us.



  • Designed for an inspection area of 240x210mm

  • Autofocus

  • Up to 24 Mp resolution photos

  • Remotely controlled camera

  • Seamless inspection process

  • Inexpensive compared to regular AOI commercial solutions

  • LED lighting systems: designed to get rid of surface component shadows

  • Capable of inspecting a board every 3-4 seconds

Using a smartphone to check the PCBAs integrity

You will need good camera smartphone. The better the camera, the better the visual inspection detection. Almost any mid-range, not too old, smartphone will work.  Place your board on a light color surface. The photo has to be taken from a distance of approximately 20-22 cms (8-9 inches).


We strongly recommend using a tripod so the smartphone will stay steady during the process. 

1) Get your smartphone

2) The reference board 

First, you will need to set a reference board. An example of what the neural network will use to compare all the PCBAs with. A "golden sample" This reference board will be a pre-checked and working condition PCB. 


  • Do not use digital zoom

  • Center the board/wafer in the photo

  • Avoid to all cost light reflections on the board. Uniform lighting conditions will bring the best inspection results.  Keep in mind that the neural network will not see what a technician can not see on the photo. 

Avoid casting shadows

That will reduce the visual inspection capability to detect faults

Keep away direct light sources from the PCB.

They will create reflections on its surface

A good ambient light 

and a detailed photo will drastically improve the detection capabilities

3) Take photos of the boards to inspect

Now, once the reference board has been set. Take photos of the boards to inspect.

A perfect alignment or placing the board always in the same position is not mandatory. The neural network will take care of correcting perspective and orientation but try not to place the boards too far from reference board original location 


  • Do no cast shadows with your hand or any other object

  • Blurry photos will provide poor results

 Using the WebAPP to inspect the boards

The WebApp allows you, in a fast way, to inspect the board using a simple user interface. Watch the video below to know how to use it.


You will need a USER ID in order to get access to the inspection tool. Contact us if you are interested. Once you get one, you are set to go.

How to use the WebApp

NOTE:   Either you are running the WebAPP in a tablet/smartphone or Computer, the processing time and results will be the same.


We recommend using a computer to run the WebApp as that way you can connect the smartphone via USB, and drag and drop the photos taken directly to the WebAPP (as shown in the video above). To do that, connect the smartphone to your computer using a USB cable, set the smartphone in "storage mode" and ready to take photos. Open its DCIM/Camera folder to access to the latest photo taken. Then, just drag and drop them into the WebApp.

If you are using an Android device, once connected to the computer you will be prompted for an option. 

       > Select TRANSFER FILES

Then, from the computer, go to the CAMERA´s photos folder. There you will find the latest photos taken. Once there, just drag and drop the photos of the boards to the WebApp

4) Wait few seconds to get results


Try it for FREE

WebApp user inteface