How AI influences the CELUS software development
With increasing use of AI, more and more start-ups are focusing on innovative ideas that will lift technology development to the next level. CELUS found a unique way to apply Artificial Intelligence and Machine Learning in the field of automating electronics engineering, we call it augmented model-based hardware development.
One can say that artificial intelligence is a term that inspires both excitement and confusion in people’s minds. Nowadays, AI is an important part of our lives, it can be seen in many every day products - from intelligent personal assistants in smartphones, driverless cars, special purpose robots for emailing, navigation, web search and social media or even in some of the fields like marketing and HR.
AI at CELUS
Artificial intelligence makes it possible for machines to learn from experience, adjust to new inputs and perform human-like tasks. Currently, it enables CELUS to integrate information more effectively, analyze data, and use the resulting insights to improve decision-making for our automated electronics engineering. Most AI examples, that we hear about today rely heavily on deep learning and natural language processing. Using these technologies, models can be trained to accomplish specific tasks by processing large amounts of data and recognizing patterns in that data. That is also an important use-case of AI for CELUS.
When electronic designs in the form of schematics and PCB layouts are uploaded into the CELUS Engineering Platform, our AI algorithms run analysis to interpret the information available on these files and make predictions. The predicted information is used, together with information provided by the user, to document internally for example functionalities, which can be used on the design process later on. To be able to make these predictions, a lot of training data went through our neural network model until we reached the high success rate we have now. However, our algorithm is still improving: Corrections made by users on predicted information are used to re-train the model, improving accuracy over time.
In order to hone our automation algorithms, there are a lot of design rules to be taken into consideration. This means there is a huge solution space of different configurations that need to be analyzed. These analyses heavily influence the output quality and computational effort of our model-based design algorithms. CELUS uses Supervised Learning to estimate those needed configurations and other parameters for converting the input specification of our customers into actual human-like hardware designs for electronics. Find more about differences and characteristics of machine learning and supervised learning in detail in this article.
We will see many more news and inventions in the AI domain, the world is changing rapidly, and a new era of technology is arising.