Issue link: https://iconnect007.uberflip.com/i/1398328
14 SMT007 MAGAZINE I AUGUST 2021 figurations, specific machines and capabilities, throughput indicators, etc. Even digital twins of machines themselves can be included. In the- ory, an EMS or fab could say, "Here is my dig- ital twin for manufacturing. I'm going to send it to my design customer." Now the problem: this information is proprietary, and it has asso- ciated intellectual property. ere is always the risk when exchanging data such as that which contains IP, because if that data gets into the wrong hands, it creates all kinds of competitive and security problems. How do you then secure that data? We're looking at some rather interesting new dig- ital techniques out there. One is called verifi- able credentials. For example, let's say a young person tries to go into a bar. In the U.S., they need to be 21 years old. e person turns up at the door, and is told, "You need to be 21." Now, the person doesn't want to say their age, and certainly not disclose their date of birth. ey don't want to share any private informa- tion at all. Let's say, however, that this person has a secure digital twin of themselves some- where in the cloud. e requirement to enter the bar is given to the digital twin, likely in the form of a question. Can this person enter the bar? Yes or no? And we hope the answer will come back, "Yes." In this interaction, no pri- vate information has been exchanged, other than the fact that this person wants to go into the bar and is thinking that they are eligible, which we knew already. No actual informa- tion whatsoever has been divulged between the parties. e person's information has been kept private. What we're working on now is a set of these verifiable credentials that provides needed information about design and manufacturing, but which does not expose private information that needs to be protected. is next step of the digital twin evolution is to work out how pri- vacy can be assured such that information can be exchanged, and in theory, used to the fullest extent, without this risk of loss of privacy. Matties: From the designer's desk, if they're doing this step-by-step through a process, they could go through each step and validate their design decision. Ford: at's a great example because there would be solutions based on the digital twin where the designer says, "I want to put this particular chip here. Here's my part number, here's my location and orientation." He then presses a button, or gets automatically, a deter- mination of manufacturability. is would work by referencing the manufacturing site's digital twins, using verifiable credentials to ask something that could be summarized as, "Is this okay?" e solution would be working on the private data, understanding whether this was okay or not, and simply coming back with the answer to say, "You're good," or not. Matties: If they did that for every step of the manufacturing process, their digital twin becomes a "live" DFM tool. Ford: Exactly. It's an interesting application of the digital twin, because it is an ability for peo- ple to be able to run these private functions, within private solutions in a way that shares data interoperably. It works both ways, design to manufacturing and back, and even out into the market.