Monday, July 13, 2015

Driving Automation ... Sooner?

Automatic Cars Or Distracted Drivers: We Need Automation Sooner, Not Later argues professor Don Norman.  Technology is moving that way. It's not a matter of if, it's a matter of when.

In 2005 Stanford's "Stanley" a driverless VW SUV won the DARPA award for negotiating a grueling 132 mile off-road race in the California/Devada desert.

By 2010 Google had several driverless Prius cars plowing streets in California practically accident free.

In 2015 the Google car is unveiled with no steering wheel and a top speed of 25 mph.

When will the ability of driverless vehicles meet and exceed the ability of drivers? Driver distraction and age effects are helping technology by raising risk which is better controlled in driverless cars.



Saturday, July 4, 2015

Driverless Cars: Does Google Have Answers?

Lou Frenzel who writes and teaches about electronics and communications recently opined Just Say No to the Driverless Car. His article includes 15 questions and the 16th one is Does Google Have Answers?

I have an answer for Google, but first, Frenzel's 15 questions for driverless cars:

  • Can driverless cars operate as safely at night as they do during the day?
  • Can driverless cars handle rain, fog, and snow?
  • Once a driverless car gets you to your destination, can it find a parking place in a parking garage or on the street? Can it navigate your garage?
  • What if you want to go for a casual Sunday drive with no particular destination? Does the car have a “browser” that lets it just wander in a highlighted area, or what? Or will that even be allowed?
  • Will drivers get frustrated in navigating around slower, more-cautious driverless cars?
  • Can a driverless car ever make a left-hand turn across traffic, make a right turn on red, or merge into heavy traffic? In many cases, some risk is necessary to make any progress.
  • Can a driverless car find a toll lane, navigate road construction, or find a detour?
  • Can a driverless car operate in New York City traffic?
  • Will the driverless car really improve a person’s productivity if relieved of driving duties, as proponents claim?
  • Will there be an increase in the incidence of motion sickness in non-drivers, as some expect?
  • Whose insurance company pays in case of an accident?
  • Will driverless cars really reduce deaths and injuries? Supporters say yes, but this has not been proven.
  • Will driverless technology come to 18-wheelers? Scary thought.
  • Will driverless cars really be affordable, or just too expensive like electrics?
  • Why not just apply all the good technology to regular cars or make a driverless mode an option?

  • The last three questions are easy to answer.
    18-wheelers? Yes, for testing purposes: Daimler’s Driverless 18-Wheelers Approved to Cruise Nevada’s Highways
    Affordable? No. Driverless cars: 15 things you need to know. System costs start at $70,000 (plus the car.)
    Optional driverless function? Yes. It has started with intelligent cruise control, lane keeping, and other piece-meal components available now in mid-range priced vehicles and above.

    So, does Google have answers to all these questions? Google rarely published or debates the merits and demerits of their current state of the art.  My guestimate is Yes, a Google Car can do all of the above with over 99% reliability if, and it's a big if, all traffic is limited to 30 mph or less.




    Tuesday, June 9, 2015

    Understanding Public Transportation Policy

    This is a eureka moment.  The following is the only rule one has to know to understand public transportation policy in the US and first world socialist countries.

    "It must always be remembered how cost-effectiveness works in the public sector: the cost is the benefit." --Thomas Rubin

    It is finally distilled!  The Cost is the Benefit.

    A region or a nation prospers when benefits outweigh costs for all public projects.  If Benefits are $$$$ and Costs are $$, then the benefit/cost ratio is 2.  That's a good project. It yields $2 of benefits for every $1 spent to build it!

    But look at Honolulu's rail where the benefits are $ and the costs are $$$$$$$$$$. The benefit/cost ratio is less than 0.1 and the entire public sector and political elite are strongly in favor. Why? Because, following Rubin's Rule, its enormous cost of $6 Billion and counting is the benefit!

    The alternatives analysis eliminated a $2 Billion light rail and a $1.5 Billion HOT lanes.  Not enough Cost... excuse me, not enough Benefit.

    Unfortunately this is a certain indicator that a society has began its Roman Empire decay.