Issue link: https://iconnect007.uberflip.com/i/1419905
36 PCB007 MAGAZINE I OCTOBER 2021 date, conventional direct imag- ing systems make modest scal- ing or feature manipulations in an effort to "fit" the image to the panel. e problem is that fabrica- tion materials and processes are colluding to require more than "modest" adjustments to scale to meet requirements. Consider that we now have panel designs with multiple lamination cycles of exotic material builds, then panels are mechanically drilled with positional and registration error and/or laser drilled with completely different error levels. It makes for a very complex picture when the end user is looking for tight annular rings and ever smaller feature sizes. DART is designed to be a path to solve the most difficult manufacturing problems we face out in the future. Feature sizes will only get smaller, drilling denser, and materials will be more exotic; we know this is a certainty. At Miva, we have a solution path to handle the most difficult of these issues. Existing Miva users can use DART today to transform the digital photolithography pro- cess into a factory floor-level tool to provide automatic compensations for variations in pro- cesses, such as develop, etch, or plating, to ensure that feature sizes and processes are per- forming at optimal levels. In late 2022, Miva will be introducing its full vision for Industry 4.0 with its patented DART solution. at is, an integrated module to mea- sure the as-built location of all designated panel features in-line with the next-gen direct imag- ing system, to allow DART's artificial intelli- gence kernel to adapt all panel features with greater precision, what we know today as "scal- ing." When coupled with Miva's post-develop/ etch/plating feature size control, the Miva digi- tal photolithography system turns the Industry 4.0 concept into reality with a real-time control center with feedback from all processes. Johnson: Am I correct in hear- ing that DART is the start of a smart factory? Hogan: DART is a smart factory. In the previous machine ver- sions, we perform small parts of DART, but we were limited by computing power—the nat- ural constraints of the tech- nology we were using. By cou- pling our next-gen light engine and new processing power with the DART artificial intelligence kernel, we're able to take a wider array of data inputs into the equation when we do our dig- ital adjustments for the feature size, position, and other factors. Controlling feature size and scale accu- racy in direct imaging is more than meeting your customer spec. Some applications don't require hyper-precision, but DART is also about opening the process window, and this is where some customers must really focus—the idea of opening a process window and con- trolling your factory in an automated way or under a reasonable paradigm. By opening the process window, the etch and strip process gets much better and can produce more with the same equipment set. If your features are more repeatable, you can develop finer fea- tures to a higher yield. If you can plate your features or come back and do a secondary plating operation with confidence that every- thing is going to be accurate, the investments that are made elsewhere in the factory are eas- ier to justify, because the factory can handle increasingly higher technology. It's a challenge for many of our customers to make this huge investment in direct imaging in general, whether it's ours or anybody else's. It's usually one of the most expensive pieces in the building. But our new DART technol- ogy, coupled with the next-gen engine, really should be having our customers re-evaluate how they think of direct imaging. Brendan Hogan