Data Analytics for Engineering Projects - EEs Talk Tech #23

How to get good, accessible data NEW Podcast channel ► http://bit.ly/KeysPodcastSubscribe Click to subscribe! ► http://bit.ly/Scopes_Sub ◄

more at http://www.EEsTalkTech.com

Twitter: @DanielBogdanoff https://twitter.com/DanielBogdanoff

Check out the EEs Talk Tech podcast: https://www.youtube.com/playlist?list=PLzHyxysSubUnAMeCIi2-S0Vm7YtSAGqx_

1:10 - Who uses/needs data analytics?

2:00 - for a hobbyist in the garage, they probably still have lots of data. But for a one-person team it's much easier manage data.

Medium and large size teams have a LOT of data from prototypes, tests, etc.

3:25 - The best teams manage their data efficiently. They can make quick, informed decisions.

4:25 - A manager told Brad, "I would rather re-make the measurements because I don't trust the data that we have."

6:00 - Separate your properties from your measurements. Another way to say it: separate your data from your metadata.

Separating data from our production lines, prototype units, etc. helps us make good engineering decisions at Keysight.

9:30 - Data analytics helps for things like analyzing simulation data before tape out of a chip.

10:30 - It's common to have multiple IT people managing data for a specific project.

11:00 - Engineering companies need a data analytics tool that is data and domain agnostic.

11:45 - Many teams have an engineer or two that manage their team's data. Many times it's the team lead. They tend to get buried in data analytics instead of focusing on engineering and analysis work.

It's a bad investment to have engineers doing IT work.

14:00 - A lot of high speed serial standards have workshops and plugfests. They test their products to make sure they are interoperable with other systems.

15:30 - We wil get industry-wide data and let engineers see how their project stacks up against the industry as a whole.

16:45 - On the design side, it's important to see how the design team's simulation results stack up against the validation team's empirical results.

18:00 - Data analytics is also a big deal for manufacturing. About 10% of Keysight's R&D verification tests make it to manufacturing. Manufacturing lines also have a different set of data and metrics that they care about.

19:00 - Do people get hired/fired based on data? In one situation, there was a lack of data being shared that ended up costing the company over $1M and 6 months of time-to-market.

#electricalengineeringpodcast #engineeringpodcast #bigdata #dataanalytics #analytics #engineering #electricalengineering #electronics #electricalengineering ... https://www.youtube.com/watch?v=atsZlEx0WTg