Image: © metamorworks / Adobe Stock
There is much talk today in the manufacturing industry of the “Fourth Industrial Revolution” and how it’s going to transform businesses. At the heart of this idea are two main concepts: Automation and Data. Automation referring to robotics, conveyor systems, and CoBots, while Data refers to analytics, artificial intelligence, and “big data”.
Automation is perhaps the better understood of the two. Robotics is a growing field, and whole automated production lines are visually impressive. Data on the other hand is not always so visible but is just as important. There is a growing market of low-cost sensors, countless “internet of things” interfaces, and the increasing sound of servers humming away in the background of many factories.
What is interesting now, is the varying amount of importance other industries put on data. In tech there is huge emphasis put on every single cell of data available. Finance is another industry that is hungry for any information they can get.
Many Manufacturers do not yet seem to put this emphasis on data in the same way. Many engineers are happy with rough estimates, manually-worked spreadsheets and the occasional stopwatch to check if everything is going well. This is perhaps for three reasons:
– Most factories still work like this
– There is little knowledge about how far (big) data analytics has advanced
– Very few have the skills to take large amounts of manufacturing data and make use of it
It has been interesting working with manufacturers in this “data” environment, while reading data analytics/processing, storage guides and articles targeted towards those in digital marketing, finance, and other technological industries. Learning what is now possible with the latest tools and seeing what manufacturers are doing today, shows significant opportunities.
What would this improvement look like?
The simple answer is every question or query you could have about your facility could be answered in a few clicks:
– What is the current bottleneck of the process?
– Do we need 4 or 5 workers on line 2 this week?
– How likely is this machine to breakdown in the next 24 hours?
– Can we safely reduce our holding stock without issues?
Information like this allows decision makers to justify their choices, respond quickly and correctly to challenges, and foresee issues in the future through simulations.
There is also a growing trend for increased customisation of products. Customers will normally expect several choices of colours at the very least, and thousands of different variations from the market leaders.
Good data flows, clear visualisations of information, and concise KPI tracking are all key to staying competitive in this market. Having the required interface and one central and accessible location means advanced analytical tools can be used to optimise the process. This includes things like discrete event simulation and machine learning algorithms.
So how can businesses achieve this?
At HSSMI we’ve had a lot of success using tools like Knime and Alteryx, which are fantastic node based analytical tools that can dramatically speed up data processing. For visualisation, Tableau, Qlikview and PowerBI are all market leaders in creating dashboards and KPI visualisations, specified to a customer’s needs.
Good Relational Database design and optimisation can help make the most of regular SQL servers and allow the storage of different types/sources of data into one location. Finally, for simulation we’ve been looking into how we can use this automatically captured data to feed directly into the simulation, saving weeks of work in validation/verification of models, using anylogic. These models can then be used to predict problems, plan or optimise flows, and help operators focus their efforts where it’ll have the most effect. Through tools like these we hope to be able to help manufacturers increase their Industry 4.0 readiness, and ultimately boost their productivity and achieve their wider strategic objectives.