Artificial Intelligence (AI) has been a number one topic for a while now. It is supposed to solve any type of problem and permanently revolutionise our daily life. In many cases this is unrealistic utopia. However, in some areas AI is not overestimated but underestimated. Especially when it comes to big data, AI can accomplish a lot more than just increasing efficiency in data processing.
Typical visions of Artificial Intelligence are mostly about humanoid robots, autonomous or even flying cars and voice-operated assistants that provide support in daily life issues. While some of this is still far-off fiction, nowadays state-of-the-art technology comes quite close to some of these futuristic ideas. Still, public discussion rarely focusses on areas in which AI could really make a difference. This applies for industrial use in particular. Eike Michel, Director Research and Development for Aucotec, clarifies: “The major strengths of Artificial Intelligence are pattern detection and autonomous optimisation, whereas real creativity and intelligence are still far away.”
Making piles of data usable
Especially in the area of data analysis, AI can accomplish a whole lot. It “understands” the data it is processing and can therefore draw connections to other data, which is not possible with standard statistical methods. Moreover, the use of AI hugely expands the amount of analysable data, as it allows the evaluation of bigger and more unstructured sets of data. Data that exists in a form that was not set up for analysis does not challenge Artificial Intelligence. Improved evaluation of data is more than just increased efficiency. “With the help of AI, we can work with data that was simply not usable in this way before,” Michel says.
Shortened sales cycle in plant manufacturing
A great challenge for companies in plant engineering is preparation of quotations, as many parameters need to be considered. No plant equals the other, therefore offers cannot be simply copied. In addition, most plant manufacturers own a huge database of completed projects. In the past, this data was rather useless for preparing quotations as it is bulky and not compiled for future analysis. Here, AI can successfully work as an interface that translates an unstructured query using human language into a data inquiry. A request like “show all plants with the features X and Y from period Z” is just one example how to use it. Based on the results, it is possible to prepare quotations with similar or equal specifications faster, and thus shorten the sales cycle.
AI adds value
With analysing and controlling gigantic amounts of data, AI cannot only answer requests about existing data but also make decisions and precise predictions for the future. Industry can make great use of these features. Predictive Maintenance is only one of several applications. Next to structuring data, AI also brings additional value that could not be created with conventional methods of analysis. “Thanks to the massive improvement in dealing with data, we cannot only handle existing processes more efficiently, but there are also new possibilities opening up that conventional computing power could not cope with,” Michel explains.