Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
Check out the new SourceForge HTML5 internet speed test! No Flash necessary and runs on all devices. ×
Transportation

It's Harder To Get an Uber or Lyft If You're Black, Study Says (time.com) 476

Black riders have to wait "significantly longer" for their Uber cabs and experience "double" the cancellation rates of white passengers, according to a new study published by the National Bureau of Economic Research. The study, which also observed a similar pattern among Lyft drivers, claims it has found "significant evidence of racial discrimination" in ride-hailing services based on a pair of experiments in Seattle and Boston. From a report on Time: Researchers pulled data from more than 1,400 field tests conducted using mostly Uber and Lyft, but also traditional taxi services. The findings in Boston and Seattle showed evidence of discrimination that manifested in either longer waits or a higher likelihood for cancellation. In Seattle, African-American UberX users on average waited 5 minutes and 15 seconds for pick-ups -- roughly 30% longer than white riders, who waited 4 minutes on average. Lyft users did not experience a significant difference during the experiment. When the research assistants switched between using white-sounding and African-American-sounding names, they did not find a significant increase in their wait times. But the overall rates at which drivers canceled the ride after it was assigned to them was more than one in 10 for riders with black-sounding names, roughly double than for riders with white-sounding names.

Comment Re:Not twice as safe I feel (Score 1) 379

In the UK nationaly there is one fatality every 174 million miles so about 1/3 more distance covered on all types of roads and conditions in mostly much older vehicles than a brand new Tesla. I'm sure I've been in vehicles with worse drivers but I haven't enjoyed it and I try to never repeat the experience.

Comment Re:This guy has a better idea (Score 1) 221

Hopefully this isn't going to remain the case for much longer as regenerative braking systems do seem like a low hanging fruit with the potential for extremely high efficiency. I'm not sure if the F1 KERS solutions employ anything with more general promise than flywheels but it is bound to be a fairly active area of research by others. Toyota and Mazda have tried super capacitor solutions and I don't think it will be very much longer until it is normal to recover almost all the available potential energy when braking in any conditions.

Comment Re:Please be an Onion link please be an Onion link (Score 1) 221

Actually the 11-16% converting to 2.3kW in the Tesla doesn't really make sense as a conclusion. Also both those links actually seem to point to the significance of rolling resistance as they include the ICE and the combustion inefficiency dominates everything else.
Looking at the fueleconomy.gov link and only considering the useful output power which is what I was really imagining above; probably less than 55hp for the typical car at cruise and ~22kW for the Tesla. Then while maintaining a cruise the useful power is only split between:
  • Rolling Resistance = 7-11%
  • Air resistance = 11-16%

Depending on speed either drag or rolling resistance are going to dominate the power required. I really had never considered rolling resistance to be of great importance to a car despite my experiences on a bicycle.

Comment Re:Please be an Onion link please be an Onion link (Score 1) 221

Thanks, my gut tells me the rolling resistance figures given there are too high. The 11-16% aerodynamic drag given in your two links matches up nicely with the 2.3kW which I estimated was required to overcome drag in the Tesla.
I've got an EcoBoost Fiesta and during the winter here (Canada) the pressure gauges on most of the air pumps at the gas stations seize in the cold and so are practically useless. When it was finally warm enough to accurately check the pressure in my tires they were all around 30-32psi. After inflating them back to the recommended 38psi I noticed my economy improved from 6.5 l/100km to 5.5 l/100km although admittedly the average temperatures had also warmed up from around -15 to -5C. Even so it was enough of an improvement that I was surprised and will be checking tire pressures more frequently in future.

Comment Re:Please be an Onion link please be an Onion link (Score 1) 221

A typical rolling resistance of a car on asphalt is 441 N. So the power at 60 mph (or 26.8 m/s) is

441N x 26.8 m/s = 11.8 kW

Using wikipedia figures the Tesla seems to require approximately 22kW at 60mph but it is extremely low drag and probably has lower rolling resistance than the typical value above. However I'd bet most cars are using 40kW (~55hp) or less to cruise at 60 resulting in the rolling resistance being responsible for 25-50% of the power requirements. I'm not convinced it would be worth it in many cases but recovering 10% of the energy could provide 1kW, the same as a 70A alternator.

Comment Re:Now if I could just type... (Score 1) 165

I assume the whole point is that it is easy to move the brains around with you. A MHL cable instead of HDMI would provide power along with a more flexible cable. Putting the brains and the input interface to a touch based OS in an easily portable devices could be a much more flexible solution than the Chromecast and its ilk.

Comment Re:If you don't want to upgrade your box (Score 1) 100

I've built a a handful web servers hosting live HLS streams for PEG and hospitality customers and RAM disks are a very simple solution that works great for me. It doesn't take much memory to store just ~30 seconds of a hundred different streams, the encoders can use webdav to push the streams onto the server and Nginx (but probably almost any other webserver) can easily serve 10's of Gbps on the cheapest of the E3 Xeons.

I can't think of a cheaper and easier solution than a RAM disk for this particular application.

I'm trying to do a similar thing on the RPi to build a CCTV system. Live streams straight from RAM but to then use the hardware calculated motion vectors to trigger recording onto an SD card. Unfortunately I'm currently stuck trying to find a tool that can perform the HLS packaging of the segmented elementary streams the RPi camera can produce.

Slashdot Top Deals

Q: How many IBM CPU's does it take to execute a job? A: Four; three to hold it down, and one to rip its head off.

Working...