Problems of implementation of artificial intelligence in the industry and the methods for its solution
The implementation of artificial intelligence helps companies reduce expenses. The automation of processes offers the possibility of reducing the workforce and, along with this, the percentage of defective product. These technologies are used by IT companies and banks internationally. Ekaterina Lyapina, artificial intelligence expert at Zyfra, and Ilya Izmailov, transformation director, explained the cause of the late emergence of technology in heavy industry and offered their opinions.
Technology implementation problems in industry
The application of artificial intelligence (AI) can increase productivity by 14%. The new technologies are implemented in the commercial sphere and the banking sector, but not in heavy industry. This sector is traditionally considered problematic for any innovation and lags far behind in terms of the number of IT solutions implemented.
The main difficulty is the high price of error. Large industrial companies cannot afford to test AI until they are sure of its efficiency. For this, it is necessary to hire new specialists, carry out studies and invest large amounts of money. The lack of data and its low quality do not allow to implement AI: it takes a lot of analysis and hiring new employees. All these difficulties slow down the development process.
Heavy industry is used to long-term investments, 10-20 years, and that do not involve sudden changes in the negotiation processes or implementation of new technologies. To start a new company, huge investments are required. Next, you need to fine-tune the process in order to increase productivity, and that’s where AI can help.
Until now, the industrial sector is not morally prepared for it. Only 31% of companies in Britain believe that it is time to start implementing artificial intelligence. This shows the slowness of the restructuring process in the industrial sector.
Very specialized and small tasks are easy to solve with the help of AI, but large tasks require working together with other companies. Large companies in heavy industry are still not talking about the full implementation of artificial intelligence in their operations and only evaluate small projects. With the help of these, the potential benefits of artificial intelligence can be evaluated and staff trained.
The implementation of AI to optimize processes offers fast and measurable results, which is why the main new projects emerge precisely in this area. Prudence and customs slow down development and do not allow the industrial sector to go to another level of production with the implementation of artificial intelligence technologies.
Comprehensive solutions in the field of AI
The pioneers already implement cutting-edge technologies in production. One of these companies is Tulip. It is a startup in the field of interfaces, on the basis of which the IIoT platform with intelligent applications is developed. Fanuc Corporation, the manufacturer of automation systems, also uses AI: with its help this Japanese company teaches robots. Freeport-McMoran is the lowest cost copper manufacturer and also employs modern technologies. At a smart quarry in Arizona, USA, they successfully tested the possibilities of AI.
Artificial intelligence technologies continue to develop and are ready to offer their services. One of the examples is the predictive analysis system in discrete industry. With the help of this system, it is possible to predict the wear of a machine or warn about the timely replacement. These practices will reduce risks from equipment defects and will also facilitate maintenance.
Another example is computing solutions for virtual consumption measurement that help measure gas, steam or fluid consumption with the help of AI. The main advantage of this technology is the high precision of measurements (95%) impossible for a conventional flowmeter.
This technology is very suitable for use in the oil industry, for example. A virtual flow meter allows to offer extraction indicators in real time, which has a very positive effect on the entire company. This device, like others based on AI, requires constant updating and data collection for self-learning. This will allow you to increase the precision of the measurements.
The emergence of new artificial intelligence technologies can significantly change investors’ views of these businesses and become an impetus to create other projects.
Why is it not implemented yet?
The cause is in the administration and the traditions. Not all companies in the industrial sector have their own IT department that is fully involved in the business management process. Unlike other industries, managers pay more attention to finances and long-term profit. Retail and e-commerce companies began to involve IT specialists much earlier, due to the characteristics of the sector and to get results from their investments sooner.
The participation of business representatives in projects with the use of AI plays a key role in obtaining benefits. Positive results depend on timely support from management. The market can grow faster, even in the conditions of a small base. The following factors have a positive influence:
- digital transformation in the industrial sector is in fashion;
- positive experiences of implementation in other companies;
- the positive results of the implementation of pilot projects and the statistically significant economic effects;
- the comprehensive implementation of successful pilot projects;
- the availability of state programs to support initiatives in this area;
- the availability of personnel and companies for transformation.
At the moment the lack of interest of heavy industry towards AI is not very reflected in the business indicators, however, some improvements could be achieved. A deep knowledge and understanding of the industry’s own business processes is required. For this they invite specialists in consulting services, but they do not risk offering an inefficient strategy that costs the company a lot of money.
In large companies in the industrial sector, IT department employees can already offer recommendations to achieve greater productivity. They use existing data and models to create forecasts and minimize errors.
How to start?
When the company finally decides to implement technologies based on artificial intelligence, it is necessary to develop the experimentation methodology, evaluate the availability of experts in the field and calculate the expenses. At this stage there are already problems due to lack of knowledge and experience.
The option of collaborating with a large IT company can be much more attractive. Due to the high hopes placed on information technology, it is necessary to start the implementation of artificial intelligence in a small project: train staff, prepare teams, evaluate how this will influence revenue.
The most important solution will be the creation of the position of director of digital transformation. This person will develop the strategy and draw up the roadmap for the implementation of information technologies. To achieve the desired results, the IT department needs to do a proper job. The initiative must flow from the top down, and its implementation requires a strong leader for the role of director of digital transformation. Your task is to organize a coordinated work between the AI and the human being.