I would note that, in recent the raw number of crashes and deaths nationwide has decreased and that is a good thing. There is anyeve evidence based on the best the problemta that is increasing or decreasing as a share of all crashes. If we look or use conservative estimates that we have across of of the more solid studies the proportion of crashes that involved drowsy drivers and wely that to 2012 data, estimate over 5000 were killed in crashes that involved a drowsy driver. You mentioned multiple you look atnd alcoholimpaired driving. I am wondering if you can use the same technique on that data to look at fatiguerelated crashes. If so, or if not, what would your opinion be . I believe it could be done. I would only do that with a couple of the data from studies that i referred to in my presentation. The National Motor vehicle Crash Causation survey and automotive sampling system. Researchers,ies, independent of the police, make the determination of whether a driver is drowsy. They di