Members 
United Kingdom
Sheet Metal Solutions
Sheet Metal Solutions
Contributed by: Luis Galo Corzo, AI Manager at Lantek
What’s more, like all technology, it can have a certain rate of error because it’s based on algorithms seeking the best possible result, which, in some cases, may not always be the best, making professional supervision necessary in order to avoid or control possible errors –something else that shows how AI will never be able to replace humans, at least in the short-medium term. But will there come a time when the algorithm has learned so much that human intervention is no longer necessary? Maybe in the distant future, but I don’t think it’s likely for complex processes, right now they learn from what humans do. Time goes by and new versions, new processes and technologies are continuously emerging. While there are new options to choose from, AI and people will have to advance side by side.
In this respect, we’ll soon see how it will replace tasks and reposition jobs, but just like in previous revolutions, new jobs will emerge to provide support for the new needs that arise, such as artificial systems trainer, AI maintenance technicians, etc. Such are the dynamics of industrial transformation where, as in the past, there is a transition period, which we are currently experiencing with the development of Industry 4.0. and hyperconnectivity
What can AI help us with right now in the Sheet Metal world? Let’s take a look at some examples:
- Optimizing scrap. One of the handicaps in the production of metal parts is the enormous amount of scrap generated. AI makes it possible to optimize the placement and cutting of the parts to reduce the generation of material waste as much as possible, with the consequent reduction in production costs. The less scrap created, the lower the cost.
- Anomaly detection. Suppose that we are going to manufacture a part that involves four processes: cutting, punching, painting and welding. If one of these processes encounters a problem, it is identified and the program sends an alert. At the moment, a technician is required to resolve the incident. There are simple processes that are already automated where this is not necessary, but the focus is that, with machine learning, there will come a day where intervention will only be necessary in special circumstances. There are cases where the optimal solution isn’t the right one, in these situations humans are very flexible, we’re capable of making bad decisions on purpose at specific times based on very little data to adjust a certain process.
- Workload distribution. AI fully visualizes all of the data that flow through a plant during the production process so that, if one cutting machine stops, it automatically diverts production to another preventing manufacturing interruptions.
- Automation of quotes. With an AI tool, quotes can be made in a much quicker and more efficient way.
- Order prediction. Using historical data, we can analyze patterns of behavior in demand, view trends and, subsequently, anticipate the evolution of consumption, both upwards and even downwards. We can also make predictions for purchasing material, staff…
Ultimately, the end goal is to automate processes to streamline production and allow people to make better and faster decisions. With the help of AI, we can accelerate decision making in production processes and automate simple manual operations.
The aim, therefore, is to try to improve industrial processes by using AI techniques applied to tools such as CAD/CAM, MES or ERP, among others, helping every day to make work easier, more efficient and productive. These tools require advanced knowledge and AI can reduce the need for detailed knowledge of processes and programs.
This intelligent software can work in both Cloud and local environments, although the cloud offers advantages such as:
- Making it easier to control and, subsequently, increased safety.
- Unprecedented access to data, as cloud environments make it possible to create much larger volumes of data than classic storage systems in physical facilities.
- Smarter decisions. It’s much easier for organizations to identify patterns and trends by connecting their data in the cloud, as AI takes data analysis to a new level by enabling far more accurate predictions.
- Easily scalable.
- It’s incredibly flexible and can be adapted to the user’s capacity demand requirements automatically.
- Consequently, with the above, costs are optimized and we no longer have to depend on conventional infrastructures and everything that they entail in terms of maintenance, updating, etc.
- It democratizes access to technology by making it accessible to any company.
The messages are clear. As an organization, whatever size, AI will help you to optimize processes and visualize factory or workshop production. It’s possible, you just have to choose to adapt and evolve, because AI will increase the efficiency of processes by reducing production costs. We mustn’t close the door to this reality that is already in the palm of our hands (cellphones), in vehicles, in homes. Digital transformation is already a reality and there’s no going back. The surge in teleworking and the increased automation of processes, something that has been accelerated by the pandemic, are simply indicators that we must be prepared for new environments and that, in AI, we have an ally.
