After each accident, damage is calculated. In most of the cases, it totals a few million euros. The facility which is halted spends majority of funds. Therefore, a failure detective, developed at the Faculty for technical sciences in Novi Sad, attracted huge attention of the local and world public.
System for monitoring of system and early failure detection called Self-Adaptive Process Monitoring System (SAMS), won at the competition for the best technical solution in 2013 in the region of Southeast Europe organized by National Instruments.
Docent PhD Milan Rapaic, assistant master Milena Petkovic and professor PhD Zoran Jelicic led the team who developed expert system, i.e. software-hardware solution which, relying on optimization algorithms and methods of artificial intelligence enables early detection of unwanted conditions and possible failures in industry.
The system is used in process industry, medicine and energy.
Prof. PhD Zoran Jelicic explains for eKapija that as of ’80, the Faculty for technical sciences opted for cooperation with economy and industrial partnership.
It is an interesting thing to mention that the Faculty for technical sciences did the entire automatic for Victoria Group which participated in the SAMS implementation.
- The solution started as the EU international project from technological development and Victoria was one of the participants –Jelicic says.
Explaining how the system works, Jelicic says that Self-Adaptive Process Monitoring System predicts unwanted situations and sends information on them.
- When one enters the section there is always a bottleneck situation which is very important for production and all things happening there cannot be measured. Operator who does not have clear data but relies on his feeling, has a great influence on that. This means you should turn that feeling into a formal algorithm.
For instance, oil is produced at 80 degrees Celsius and it can be produced at 90 as well but if it increases to 100 or falls to 70, we have a problem.
- The system is made, an algorithm based on an artificial intelligence. It is a complex software-hardware system. The production process is monitored on the computer and then a signal appears which shows that something that has not been happening before is happening. Our objective is that the system informs that something unexpected will happen a few minutes earlier. This consumes less energy, number of failures, lay-offs and accidents is directly decreased which results increased health and safety at work level. It also decreases production costs, influences more rational energy consumption and maintains product quality.
Professor PhD Zoran Jelicic says that SAMS is not a tailor-made solution but that it has a general algorithm in its basis. It can work as an individual application and as a part of the existing software systems for monitoring and management.
- The system is tested in a solar power plant in Italy, in holter devices for prevention of bad heart work as well as in a huge thermal power plant in Serbia. My idea is to present ourselves in America and to suggest to one company to offer our software along with their hardware.
Apart from the described solution, the expert system specialized for early detection of electromotor failures based on vibration signal analysis was developed.
- This system enables early notion of majority of familiar electromotor failures and so far it has been tested in over one hundred of motors of 1 kW to 3.2 MW. The main feature of the system, which comprises expert knowledge and signal analysis hardware, is that, as well as similar commercial systems, it can be used as a diagnostic tool in the maintenance process but it provides an option of integration into the existing supervision – management system which significantly develops the quality of management and system reliability.
The Excellence Center
As a result of the successful cooperation of the Section for automatics and management of the systems and the National Instruments company, by June Novi Sad should see the opening of the Excellence center for intelligent and advanced automatic management where students and engineers will have the opportunity to acquire knowledge in practice in the field of implementation of complex algorithms on industrial computers.
This will be the only center of this type in this part of Europe.
- The idea is to train engineers who went on practice and who would thus maintain their competences.
Good news from Bologna
- We have nice news from Bologna. They want to send their doctoral students to the Faculty for technical sciences in this field. Apart from that, I hope that we will take mutual part in one big project of a couple of million euros. These are serious research and development projects – Jelicic says.