Giovanni Piccolo is the Vice President of Project Management and Applications at Salvagnini. He has 36 years of experience at the company and has seen incredible change in technology during that time. Fifth Wave Manufacturing caught up with him via Zoom while he was in Salvagnini’s Italian headquarters.
FWM: What was it like when you joined Salvagnini?
GP: I joined Salvagnini after serving in the Italian army. It was 1985, and as a young engineer, I was one of the early people to write software for our machines. We were the first in the industry to use a full-blown computer. At that time, there was no such thing as a powerful Windows computer. We deployed Digital computers, PDPs [Programmed Data Processor, a minicomputer], they were the first computers for multi-tasking software. This allowed us to organize a structure to write automation software in multitask function.
You can imagine at that time the calculation power was ridiculously low compared to today. We were transferring data to a 5 MB hard disk, using 8-inch floppy disks, and writing FORTRAN—all of these scenarios. This was the third industrial revolution, which was automation. With our software, we gave the machine a brain, with all the struggle that comes with that task. We wrote a lot of software which allowed the machine to work automatically, minimizing the mechanization. The difference between mechanization and automation is that under automation, the main processes do not need any support by labor. The machines ran processes and did diagnostics.
Processes which used to be part of human activity—loading blanks, changing tools, etc.—have changed because of automation. It’s true that automation is about the electronics in the machine, but to my mind, even more important is the software. The software development, the IT development of even more compact and powerful computers allow us now to do the things that were science fiction when I started to work at Salvagnini.
FWM: How has that affected the way machines work and the way we work?
GP: In the last 50 years, there has been a change in the mentality of the industry. We are moving from “wired” logic which was not flexible; it was a machine was doing a certain sequence of actions to create some product. Now, with software comes flexibility. Since the beginning, Salvagnini designed machines that could offer a solution with a computer aid, to meet the challenges of production. Our goal was to combine excellent mechanical performance with excellent flexibility and capacity to adapt to certain functions.
In an overall sense, we are moving from big hydraulics, running on a three-phase motor, pumping a lot of oil, toward more controllable sets of controls. Substitute cylinders with a lot of oil with other devices which can be precision gear, pneumatic equipment, brushless motors, and leaving the power function only where it is needed. These factors gradually changed a machine with a lot of oil into a machine that uses the oil only where necessary.
Think about an aircraft. Landing gear are still moved by a hydraulic cylinder, but the cylinder is driven by fly-by-wire. There are no cables any more in advanced aircraft. That same thought is pervading our industry right now. Automation when I joined the company was moving to Industry 3.0, where computers are located for the use of the machine. Still, that was not enough. The machine is a living part of an organism—the organism being the company or the factory floor.
Back when I joined Salvagnini, we sold a system to General Electric. It was analog. We were receiving data via normal telephone lines. We wrote all the software and we had Digital create the commands for the machine. There was this string, this protocol that was developed, and it worked for 50 years, we were able to transfer an order from Connecticut down to Carolina, with a list of material and priority.
Everything at that time was a deep, deep dive. You created customized software, you created your own protocol, etc. With the new computers this has been moved from self-made, proprietary formation, to an open architecture situation. This became the real trend a number of years ago. All the protocols in use today were designed to be compliant with other systems that used that protocol.
FWM: A common language evolved.
GP: Yes. Perhaps you remember Esperanto, it was global language that everyone could understand and use. Our industry is becoming like that. A box can receive input and give output which are readable almost everywhere, and therefore you can use the data as you wish.
FWM: What comes after that?
GP: The next revolution will be artificial intelligence. I think the IIOT environment is gradually moving our focus on that. I don’t know how much we already live in that arena. I will say that future systems will be more independent from the direct intervention of an operator. There will be a lot of intelligence built into our systems.
FWM: Let’s move to the current functions. Nesting, for example, it really isn’t just about lasers.
GP: Nesting is a mathematic function. We have to combine objects in a defined space, with two targets in mind: first, reduce the cost of material, and second, when we nest a part, we also have to think globally. Here thinking “globally” means considering things throughout the organization.
Let’s say I want to use all the material I bought for my blanks. I make the full complement, then I have 15 percent of the material left. Here’s where the intelligence of the designer comes into play. It is only wise to use the remaining 15 percent if the surrounding technology that depends on this specific process will maintain their output. If I’m going to influence the output of the other systems, I create more or less work in progress or put strains in my process. This inclusive, global approach is a great approach.
When designing a customer system, if we have a process that produces blanks, and then the blanks need to be folded, we combine the processes. In fact, we have expanded this concept: We know this system produces goods—panels, for instance—which need to be assembled. In terms of the “global” outcome of the process, it is better to follow the assembly rather than the nesting. Therefore, we tried to give all this flexibility to our users, so they can morph the strategy of nesting according to the end goal.
FWM: Probably there are more targets to meet as time goes on.
GP: Really, we don’t have only one target. We ask, “What do you want to do?” Do you need to serve a paint line, for instance? Do you want to operate by color? Do you want to go by zip code? The nesting is part of an overall process, and now you can have a direct connection to your ERP, transforming your bill of materials (BOM) into a job list. We tried to give our customer all the power to change the behavior of system.
FWM: That’s fascinating. To do that with any level of ease of use for the operator, all the complexity lies unseen behind the interface. For the operator, it can be a simple operation like you would set your camera to give priority to the f-stop or the shutter speed.
GP: I may be a bit extreme in what I’m saying, but still, we are strong because of our main technology, our machine design. For instance, on the laser we have a lot of functions to minimize scrap and to optimize the part, we include lots of algorithms to place the part in a strategic manner. Or when we speak about our punch and shear machine, it is naturally designed to not to produce scrap, to use the maximum of the material.
There are some other technologies that have a threshold of unavoidable scrap. The machine is designed to potentially minimize so that every square inch of that material can become a part. You can win value per pound, pounds in and pounds out. We can get to very limited scrap.
The typical example in my mind is the difference between a job shop and an OEM. If you buy from a job shop, you buy time and parts per hour. The more parts you can do in a minute and the less material you use, the more money you make. In an OEM, within reason, we look at how many parts are put at the shipping gate. For an OEM it’s a good idea to start at the shipping gate and work backwards, and see where we are in the process.
Our job is not only to sell a machine that has the best nesting, etc., but also to make smart automation commonplace. You have to address the need of the customer to understand how much and what automation you have to apply. Across different companies, the targets are similar but never equal. The simple geometry of a factory can determine one type of automation versus another.
FWM: You’re designing based on future customer use?
GP: All of these things work together—overproduction, unnecessary motion, we try to keep all of these different factors in mind. Also important, automation works perfectly when it works, but the machine can have some problem. When we design an automated system, we always try to give the opportunity to use the system partially automated. We always want to have the “spare wheel,” if it can keep the system productive.
The factory is one big organism. We always take the opportunity to study a client’s workflow. We want to know what the customer is trying to achieve. That’s why our system has this capability from the mechanical point of view.
You can compare it to a box of Lego bricks. You have the main bricks, you can make a fire station or a cottage—it all depends on how you mount them, build them.
FWM: Have you ever had the experience where you have a new automation opportunity and the customer actually changed their business model or changed their factory flow to match that new capability?
GP: When we enter factories, we do look at processes. Perhaps the use case of the customer has to be changed. This is a case where you must give the maximum support to the customer. People that got used to the sequence of the job in a certain manner, won’t necessarily do the same with our machine. That is why we need to have clear targets that are agreed upon.
I have been traveling the world and sometimes I’ve seen automation that was added merely because they wanted to automate, but the solution was not efficient. For automation, success comes with flexibility, when the automation can adapt to its surrounding circumstances in a cost-effective manner.
We have gradually inserted a lot of options into our products like panel benders. When we design our panel benders, we can say that now, like the Lego brick situation, we have such a big library of opportunities that we can give to our customer. The process in this case is the betterment of a process.
FWM: The process is affected by automation and vice versa. Can you please differentiate Salvagnini’s approach in this regard? The whole setup of the software in the stream of part creation, it’s very much a process orientation, not a task orientation. It’s a global view of everything in the factory. How is it possible to differentiate that?
GP: Our approach is to offer a customer a service that is complete. What we have done and what is in our DNA is this: We have considered the machine from both the design aspect and the function aspect.
What does that mean? When you think about a machine, you have software to describe what the machine has to do, and the natural language of that is coded in the machine. We started this in the early 1990s, and we are still getting great benefit from that.
Our applications can take the geometric information from an electronic drawing, a CAD file. Our customers say that we should not have the machine show us how to do it, but what to do. That has always been our approach, we maintain flexibility. We always try to interface the machine and link the information that comes from the outside. Along the way we have developed all the interfaces to the modern CAD system.
But that is not sufficient, because a system has to produce what you sell. In the past, and still sometimes, we meet people whose sole role is to understand what has been ordered and to assign tasks to machines. All of this information is managed by ERPs. Manufacturers still work with printed paper, a bar code, and that goes to an operator.
However, things like this got solved a long time ago by integrating the Ops [both DevOps which affects development and DataOps which affects data and analytics]. The system can read the data coming from an ERP, create all the links, search for the drawing, do a nesting function, so that there is no additional input required to transform an order you might receive from the website into a product.
Our machine sometimes can seem like a 3D printer of your IT system. I write the file, I hit send, instead of a piece of paper, I will have a part at the end.
FWM: It seems like as a vendor you have to have two simultaneous outlooks: machine-based and global in the way we’re defining global here. The operator shouldn’t have to think globally if their performance is based on a step in the process. But also thinking globally, sheets come in one end of the building and parts go out the other, and you have to respect that, too.
GP: That’s a fair assessment. On the part side, we have an acronym, PART, which stands for Precision, Automation, Reliability, and Technology. The operator doesn’t have to think about all the sophistication, they think about the part. Management does think globally, and our situation now is we retrieve data from the design files and from the ERP. Now, with IIOT, the system can broadcast its status to the network, and we can manage that data.
And finally, we are now able to integrate operations that are not our core business, for instance deburring and welding. While they are not our core, neither are they detached from the whole process, and thus they are important. The part is not what comes off of one of our machines. It is what is packaged, invoiced, and sent. This is where our customers want us to contribute to management and efficiency.
For more information: https://www.salvagnini-america.com