Create value with AI - bit by bit
How can AI be useful in production and manufacturing? What is important to think about and how do you get started? On September 22, Automation Region conducted a webinar focusing on industrial AI applications.
The webinar was organized in collaboration with the company Peltarion, which is developing a graphical interface that makes it possible to create AI applications without advanced programming. The idea is to make the technology available to more people and also to create conditions for smaller companies to get started with AI.
- More and more organizations are investing in AI, but in the traditional manufacturing industry development is not so fast, says Peiman Momeni, business developer at Peltarion. Many who we come in contact with state that they would like to wait for the technology to mature and for the tools to be better suited for their specific business.
According to Peiman Momeni, AI skills are something of a shortage in the industry. However, many AI models are based on open source and that is a good thing. Companies can customize them and use their own data to start taking advantage of the technology.
- AI has great potential in, for example, error detection, predictive maintenance, automation of production and improved support for operators, says Peiman Momeni. A good way to get started is to start from a limited application, such as inspecting a machine.
When can you benefit from AI?
- When you have a lot of data that is difficult for a human to process and analyze.
- When you have different types of data sources, such as picture and sound.
- When your data consists of unstructured text and you want to identify meanings.
- When you want to automate or improve the outcome of a process performed by humans or by rule-based software.
- When you want to make predictions based on historical data from different sources.
Many actors test and evaluate different AI solutions. An example is Tesla, where the company's customers continuously collect data that step by step creates the conditions for self-driving cars. It is also clear that companies that are strong in AI - such as Facebook, Google and Amazon – can easier make moves into new industries.
In the longer term, the implementation of AI is considered a prerequisite for competitiveness, but what does it look like today? Peiman Momeni referred to a survey (Hype Cycle for Artificial Intelligence, Gartner Research 2019) which shows that only 14 percent of the companies that have invested in AI solutions state that they get value out of their investments.
- We cannot expect rapid effects in the business, says Peiman Momeni. A strategy is needed where, in our experience, it is crucial to consider AI as a core competence that must exist within one's own organization. In addition, relevant data, domain knowledge and good tools are required.
Nearly 200 participants followed the webinar led by Daniel Boqvist, Deputy Process Manager at Automation Region.