Building software systems that use data to make decisions, also called data science, is what today’s advanced technology companies are building their algorithms on. In today’s world of software products, we believe that software systems should be capable of learning from past data to make decisions, similarly to the human brain.
So how can we, as humans and software systems, make good decisions based on information from the past? The key to solve this problem is by a vast amount of information rich data. This will help us to build systems to use the enhanced understanding, to take decisions and make the most out of our knowledge.
Let’s illustrate this with an example
If you ask yourself, what colour tomatoes are when ripe, you will probably answer red. Thus, by looking at a tomato you would be able to decide if you can eat it or not. Now imagine you’ve never seen an apple before and you see an apple tree with red and green fruits hanging from it. From your experience with tomatoes, you would assume that the red ones are ready to be eaten. One day you see a tree with only green apples. Your historical learnings would tell you that they’re not ripe yet. However, after observing them for a year you realize that they never turned red, even though they fell on the ground and started rotting. At some point you realize that there are also green apples that will never turn red, but are juicy and sweet regardless. By adding a new dimension, such as type of tree, time of the year or geographical location, you increase your understanding and thus your learning of whether a fruit is ready to be eaten or not.
How can this be adopted to the manufacturing industry?
At Kreatize we deal with custom parts that our customers use to build their machines on a daily basis. How do you calculate the price of a part that has never been produced before? Mechanical engineers are able to estimate prices, by looking at technical drawings of these custom parts and use their experience and knowledge to determine the right price. This is not only done by the geometrical representation of a custom part, but also by considering other information from technical drawings.
For instance, parts with high precision tolerances have longer production times on the machine to be produced and are therefore more expensive. A mechanical engineer will be able to extract this information from a technical drawing.
With the use of Werk24’s software stack, Kreatize will be provided with systems that are capable of reading and understanding information from technical drawings. With this information gain we are capable of improving our cognitive systems, which take decisions based on the available part information.
Kreatize’s value creation
Working with Werk24, at Kreatize we can improve the data quality of our in-house CAD agnostic format for representation of manufacturing data.
Having a better data representation and better data insights, will help to improve and optimize our pricing system, manufacturability analysis and finding the right supplier for our customers.
The manufacturing industry will become more competitive as innovative companies like Werk24 tackle unique and complex challenges. By having a better understanding of our customers’ requirements, Kreatize is able to simplify the manufacturing process and digitalize the procurement process through supportive software-based systems.
Through the cooperation with Kreatize, Werk24 is able to improve their Artificial Intelligence system, by making use of new, unseen drawings and learn from errors.
Jointly, we believe the collaboration contributes to the strength of the European Manufacturing Location.