Data in Smart Manufacturing

Data in Smart Manufacturing

User Rating: 5 / 5

Star ActiveStar ActiveStar ActiveStar ActiveStar Active



Manufacturing is getting smarter with the advancement in many emerging technologies you have probably heard of many times (IoT, cloud computing, AI, digital twins, etc.). Nowadays, when the global economy stalls due to the corona pandemic, the talks about smart manufacturing increased even more. We have already shared our opinion about Industry 4.0, which is proximately related to smart manufacturing, but in this blog, we would like to specifically focus on data and also on how testing stands as one of the most valuable processes in generating manufacturing data.


Part 1. Evolution of Data.


First of all, let us briefly look at how people were collecting data throughout the history of manufacturing ( the following paper provides a brilliant overview of this topic). The general idea is that we were always collecting data one way or another. In short, during the handicraft period, data was collected in a form of a human experience. Next, during the machine age, data was collected manually and classified on machine-related (for the means of maintenance, production, etc.) and human-related data (for the means of salary structure, KPIs and etc). Furthermore, the digital age brought us very sophisticated ways of collecting and storing data (information systems, databases) resulting in an exponential growth of its amount which allowed us to create advanced intelligent manufacturing models. Today we are living in the age of big data and, in manufacturing, the big data is defined as a large amount of information gathered from multiple sources throughout the product lifecycle. Data now can be collected from MES (manufacturing execution system), assembly machines, testing equipment, internet sources (user data can be gathered from there), various external databases (e.g. scientific, governmental), and other places.


Part 2. What is Smart Manufacturing?


Furthermore, let’s answer the questions “What is smart manufacturing?”, and “What does it have to do with data?”. In general, smart manufacturing is a broad scope of activities and technologies that are intended to make manufacturing more intelligent throughout the whole value chain. There are several organizations around the world that brings manufacturers together and discuss the future of manufacturing (for the past several years they specifically focus on smart manufacturing). One of them is MESA (Manufacturing Enterprise Solutions Association) and every year they bring the white paper that shows the landscape of smart manufacturing for the current time. In the figure below you can see the framework adapted from one of their white papers (more about it in this blog) that explicitly shows what is smart manufacturing.




Answering the second question about the role of data, we can see that smart manufacturing brings together all the stages of the value chain under one roof. Therefore, there are tons of data that is constantly being processed in real-time. This gives us another way of defining smart manufacturing as “a field of manufacturing that sets its goal on collecting the data across the product lifecycle to bring manufacturing intelligence”.


Part 3. What Are the Challenges?


Despite the relatively steady hype around smart manufacturing and the development of data technologies, there are some challenges that still prevent us from seeing fully effective smart manufacturing. For example, at Exsensio ltd, we usually notice that data always comes from different sources and this data is very heterogeneous. Consequently, it is very challenging to collect this data and utilize it to its full capacity.


Furthermore, the Big Data Value Association describes challenges in smart manufacturing in their annual whitepaper. In their 2018 report, they divide the challenges into 3 main groups: the adoption of data technology in production facilities, the adoption of data technologies in the product lifecycle itself, and the adoption of data technologies throughout the value chain. In the first group, the challenges were associated with computational continuity, data distribution, and data storage. In the second group, the challenges were related to the creation of stakeholder data spaces for complex products, and also IoT-related constraints. In the 3rd group, the challenges are associated with B2B industrial platforms. This is a brief introduction of the topics and you can get more info here. Not to forget, that apart from technical challenges, there are regulatory and legal obstacles and challenges related to innovation and the creation of new business models.


Part 4: Why Exsensio ltd Cares About Data?


At Exsensio ltd we are very interested in data analytics and therefore would like to contribute to a faster transition to smart manufacturing. As we discussed in our first blog about the current state of automated hardware test, the nature of this area, and our testing framework specifically, are developed in a way that there is a need for constant interaction between testing equipment and the device under test. Moreover, the interaction is configured at a low level (the user has a more accurate control of the interfaces between the device and the equipment) during the product development stage and after the assembly. Testing itself generates a huge amount of data, and the low-level control to the interfaces can provide some insights about the device not known before.


Besides, the data collected at the end-of-line testing stage can be immensely valuable, as it is the most important indicator that the assembly of the product has gone successful, especially when the manufacturing stage is outsourced. While developing our test automation framework at Exsensio ltd, we put a strong emphasis on the remote control of the test and real-time tracing of test data. This is turned out to be helpful especially at the current times when even a company with an in-house production might need a remote control.


Another important aspect to care about the data in testing is product quality. Interestingly, when you look at the content related to smart manufacturing, there is surprisingly less information about better product quality than, for example, about efficiency, cost reduction, or smarter automation. Moreover, we can see that today, traditional inspection methods and testing software products are away from the standards of smart manufacturing. Inspection methods are still manual in some cases, and software tools are not data driven. This may also pose a question of: “What do you know about your product?”. We believe that data-driven smart manufacturing and next gen testing solutions will help us to answer these questions.




Tao, F., Qi, Q., Liu, A. and Kusiak, A., 2018. Data-driven smart manufacturing. Journal of Manufacturing Systems, 48, pp.157-169.

Kusiak, A., 2017. Smart manufacturing must embrace big data. Nature, 544(7648), pp.23-25