Issue link: https://iconnect007.uberflip.com/i/855901
22 SMT Magazine • August 2017 cycle times. Some of our machines run 100,000 or a gazillion compo- nents, but interestingly enough, we have found that in some in- stances, slowing down the equip- ment significantly improves the output and the flow. It is a com- promise. Back to your question about the goal, it could be many different aspects. It's just a matter of understanding what is it that you're trying to accomplish and what could be negatively impact- ing the goal." "We're rabid about reducing waste and we sit down, we val- ue-stream map, we do some measurements, and we see where we should focus first," says Prina. "The tough part is always how you make sure that whatever you've done sticks. We try to use mul - tiple events and things like that. Get people in- volved and then go back and just get that stan- dard work written down and then checked daily at a supervisor or lead level, and check monthly at an executive level. Did we stink? Are we still head- ed in the right direction? What's our next step? Okay , we reduced 50% of the waste, how do we get 50% out of there again?" Best Practices So, what are the best practices to consider when improving the electronic assembly pro- cess? Prina says its "measure what matters." "For years, we've picked out a number of KPIs and we just took some numbers here, took some numbers there, and didn't real- ly do much with them," he says. "We didn't get much effect for it. We built some KPI trees, decided what key processes we really want to monitor and what we're going to do with that information once we have it." "Back to fundamentals," notes Ramirez. "The theory of con- straints. There is one thing that is making your whole process slow. If you focus on that area, then you measure that area, you know that everything else flows. Listen to your employees. That guy or gal that is spending eight hours in front of that machine can prob- ably tell you more about what is causing the problem. Maybe he or she doesn't know what exactly the problem is, but by getting their in- puts, understanding under what circumstanc- es these problems happened, they typically can tell you a lot of information and help you actu- ally fix the problem. That is something that we have to keep repeating all the time to our man- agement team. The more your people know, the more you involve them, the better your results are going to be." SMT TRAINING AND EDUCATION: KEY TO IMPROVING ELECTRONICS ASSEMBLY Researchers at Stanford University and MIT have built a new chip to overcome the challenge of processing massive amount of data into useful information. The results are pub- lished in the journal Nature, by lead author Max Shulaker, an assistant professor of electrical engineering and comput- er science at MIT. Shulaker began the work as a PhD student alongside H.-S. Philip Wong and his advisor Subhasish Mitra, profes- sors of electrical engineering and computer science at Stan- ford. The team also included professors Roger Howe and Krishna Saraswat, also from Stanford. Instead of relying on silicon-based devices, the new pro- totype chip uses carbon nanotubes and resistive random- access memory (RRAM) cells. The researchers integrated over 1 million RRAM cells and 2 million carbon nanotube field-effect transistors, making the most complex nanoelec- tronic system ever made with emerging nanotechnologies. The RRAM and carbon nanotubes are built vertically over one another, making a new, dense 3-D computer architec- ture with interleaving layers of logic and memory. By insert- ing ultradense wires between these layers, this 3-D archi- tecture promises to address the communication bottleneck. The team is working to improve the underlying nano- technologies, while exploring the new 3-D computer archi- tecture. For Shulaker, the next step is working with semi- conductor company Analog Devices to develop new ver- sions of the system that take advantage of its ability to carry out sensing and data processing on the same chip. New 3-D Chip Combines Computing and Data Storage Chris Ellis