For this specific project, a live image recognition application was developed for a flexible photovoltaic production line.

• Problem

Following information from the company, which contacted Depia Automations, an analysis of the requirements and problems faced in the production process and its supervision was carried out, in collaboration with the said company and the software department of Depia Automations, specific problems were identified. The company’s managers, working on a research project to manufacture innovative flexible photovoltaics, with a state-of-the-art production line, needed a way to control all the materials that make up the photovoltaic panel during the process and at all stages of production, from which will pass. Since this feature did not exist at the given time, the accuracy and efficiency of the overall process of materials synthesis fluctuated at low levels, while errors worsened its speed and performance, as in any error the process had to be restarted . The general economy of the production process could be improved.

• Planning

The experienced programmers and developers of Depia Automations, after recording and dividing, in collaboration with the respective managers and company staff, the above identified issues, started the development planning.

The goal was to develop an application, which through a live image from a camera, which would be placed inside the production line, would check throughout for any errors. In addition to the live camera, there would also be corresponding alarms, which would operate based on the image coming from the camera, every moment, so that in the event that an error occurs, it can be detected immediately. When a modern programming language, is chosen the code development plan for each was carefully designed for each section of an experiment separately.

• Implementation

Since its development planning was completed of the software, considering all the factors, we proceeded to develop code in C#. The specific language was chosen because it is a modern, widely used and recognized, as well as easily extensible programming language. In addition to the choice of a programming language used, it was deemed necessary to configure a database adapted to the needs of the software and to connect it to the program under development in order to store all the parameters and data returned by the live camera from the production process , automatically. At the same time as the aforementioned implementations, regular visits were made to the company to familiarize the researchers and operators with the use of the used application, but also to receive feedback for its continuous improvement.

• Result

Having now an application for monitoring the production process of flexible photovoltaics, which gives the possibility to visualize the process and control, we managed to maximize the accuracy and performance of the system. The entire process is continuously monitored through the cameras and any error or deviation from the desired values ​​that occurs is now automatically observed at the exact moment it occurs.
The efficiency and safety of photovoltaic production experiments has increased significantly, since now the researchers and operators of the production stages have full control. In conclusion, the overall process became faster, more efficient, saving valuable time from the researchers and generally maximizing the economy, time and money, of the system and the company.