In your opinion, how has the Smart Factory landscape evolved over the years? What are some of the advantages of the current technological evolution?
Before industry 4.0 developments, we were sitting and waiting for our mistakes to occur, or trying to calculate our plans based on the limited data we have but the smart factory landscape is now revealing our shortcomings to us before we even don’t see them. The biggest advantage of current technological evolution is how to make the chaos you will experience with bad surprise more effective.
What according to you are some of the challenges plaguing the Smart Factory landscape and how can they be effectively mitigated?
The biggest challenges that we have experienced most prominently in Smart Factory environments.
The first is a lack of interoperability. In intelligent factory environments, when a new operation is added, communication protocols, hardware, software are the main challenges we face in intersystem interoperability. All variants of the new operation to be added make the design phase challenging to be compatible with your smart factory infrastructure. Considering the new technological developments every day, the first model you develop may not work due to the differences in the next technological development. Here’s how to effectively mitigate this challenge; standardizations in manufacturing technologies are gaining great importance in this regard. We are constantly making improvements and updates to our internal process standards. Also, our suppliers will easily adapt to their solutions with our different suppliers in modular structure (e.g. data structure, communication protocols, etc.) to produce solutions. This push increases the success of Smart Factory environments on a significant scale. In this way, multi-manufacturer systems can work with each other more easily and quickly.
The second is the difficulties in managing data growth. As all vehicles and machines operating in a smart factory environment are equipped with sensors to make them smarter, the amount of data collected from these sensors in the background increases spectacularly rapidly. Now that real-time feedback is being worked on, OT and IT teams are working hard, so to speak, to ensure data is cleared, analyzed, and inference, starting with sending data to the address to be processed smoothly. It is no longer difficult to collect data; the difficult thing is to keep up with the speed of the collected data. Instead of collecting all the unnecessary data necessary to deal with this challenge, we calculate the amounts of data that we can collect in order of priority by setting priorities. We estimate sampling rates through AI algorithms, taking into account machine and operational changes. This helps us grow, develop and keep up with our data rate in a controlled manner over certain periods, preventing us from creating unnecessary costs and workloads by collecting all meaningful, meaningless data initially.
The third big challenge is data sensitivity and security. In operational terms, data security is vital because smart factory environments are powered by a lot of machinery and manpower. Since all systems are located in a multi-network structure, all communication channels need to be managed meticulously. In this regard, our IT & OT Security teams work together to create more effective solutions by preventing the accumulation of workload on a single partition. In smart factory environments, all interactions, both physical and digital, are identified and possible vulnerabilities are evaluated. We review our security policies by testing up-to-date attack methods (including social engineering attacks) on one-to-one virtual copies of our systems during certain periods. Otherwise, each point of data sharing in smart factory environments will automatically become vulnerable to attack, so we train all our employees through the necessary training and real scenarios to address how serious operational problems they will cause.
Apart from the 3 major challenges I mentioned, of course, there are challenges that we face, but these challenges are not as big and arduous as the trendy challenges I mentioned above. For example, technical knowledge is a lack of skills, the time curve is a wide challenge and it is very difficult for us to provide the necessary workforce immediately in smart factory environments. However, with our continuous training policy developed and implemented for our employees, we increase their awareness both by updating their knowledge and by helping their personal development.
Which are a few technological trends influencing Smart Factory today? What are some of the best practices businesses should adopt today to steer ahead of competitors?
Today, one of the technological trends affecting Smart Factory is the developments in the fields of Artificial Intelligence and Computer Vision. While we used to be able to work on structured data, now operationally meaningful data can be extracted from data such as image audio. These inferences began to contribute to the results in a serious way. Also, open-source studies can take research in the field of Smart Factory to different dimensions. Today, for example, thanks to machinery and deep learning applications, an AI algorithm that you train for any process in your operations can make logical and effective recommendations that are not included in structural analyses such as time and workspace in a way that you do not expect. By calculating the network loads of your systems in your communication channels, you can detect possible anomalies in advance. By experiencing the workload and problems of your maintenance costs in your systems over time, you are prevented from encountering bad surprises by predetermining possible problems. In this way, you can analyze your production operation without interruption and additional costs.
Do you have any advice for industry veterans or budding entrepreneurs from the Smart Factory space?
My suggestions for entrepreneurs and businesses working or developing in the field of Smart Factory is to give the necessary importance to the work in the field of artificial intelligence and to support the growing open-source studies in this field. In this way, your perspective can be expanded in a very different way and unpredictable problems can be solved very easily.