Autonomous Cars: why we need to be fair in judging them

Driverless Car
(Image courtesy Institutional Investor)

After more than 20 years in the M2M/IoT space, you tend to get a lot of good stories to pass along.  One that comes to mind is when I was part of a deal that sold an in-car Field Service solution to a trucking company.  The Operations Director was definitely not in our court and had a ton of "what if" scenarios to prevent us from succeeding.  One of his main concerns was about Security, and rightly so.  He grilled us on exactly how secure the overall solution was, the chances for information theft, etc.

After a long presentation, he said, "I just don't see how secure this really is", to which one of the IT guys (who was definitely in our favour) responded, "Probably a lot more sure than us leaving papers with credit cards in a clipboard on the front seat of the car with the window open while our worker is in McDonald's".  Well, that made everyone laugh and reminded everyone that the solution we were proposing was light-years ahead of what they were using now and that it did not add any undue security risk compared to what they had now.

The same comes to mind when it comes to autonomous cars...we need to treat them on a level playing field.

No solution, even automation of cars, will prevent accidents

While I have never actually been in a car without a driver, I have been on many shuttle trains that are driver-less.  The trains seemed to run well, always stopped very precisely and I don't recall ever having a serious delay that would not also have affected a train with a driver operating it.

So, the concerns are always, "what if?"....what if a pedestrian walks out? .....what if there is a power failure? ....what if there is no Internet connectivity?

I don't want to not say that these things are important, they most certainly are.  But, two can play the “what if?” game, this time, talking about drivers and their cars.  What if the driver has a heart attack? ...what if the driver is talking on the phone?...what if the driver is a bit hung-over and has slow reaction time?

The fact is that we need to stop assuming that every driver has the reaction time of a ninja, the driving skills of a Formula 1 driver and that they’re paying attention as closely as I do when my wife talks (ok, that is not a high bar to achieve, but you get the point).

The fact is that for every one driver like this, there are millions of:

  • 18 year girls who try to sneak a text in
  • 80 year old men who mix up the gas and the brake pedals regularly
  • Soccer Moms who need to figure out how the frog got out of the tank while driving (wish this was not a true story)
  • Middle age salespeople who started their presentation (on their laptop, on the front seat) to be on time....and, I am not admitting that I ever did this, by the way.

The Bottom Line

When we are comparing the driving skills/performance of an autonomous car, we need to compare it more to the average driver, driving under normal circumstances and not what the ideal driver may do on their best day.  Inevitably, there are going to be issues in the future with driver-less cars, but it does not mean that we are worse off by using them. 

Living in a world of data, but without facts…

Quick....how many posts do you think Facebook did per hour on average back in 2015?  How many recipes, pictures of desserts and rants about Game of Thrones were posted?

If you are like me, you may be startled when you hear that it was up to 250 million per hour!

But, maybe I shouldn't be...there are over a billion active monthly users....and Instagram isn't too far behind, often hitting 100 million likes in an hour.  Poor Vine....it seems small at only a million videos being played per day.

So, why does this matter?

It means that the world produces a ton of data.  One report had us creating 2.5 Quintillion bytes of data each day.  To put that in perspective.....it would fill 10 Million Blu-Ray discs, and if you stacked them up, it would be the height of 4 Eiffel Towers.....each day!  (Thanks to VCloud news for that).

Even if this is a bit off, it still means that the world is drowning in data.  And, we have not yet seen the world of IoT fully kicking in.  I can imagine that once it does, these numbers will go from huge to not even making sense at all.

All of this data....but not a lot of facts?

I first wanted to thank the New York Times for their great Op-Ed, "The Age of Post-Truth in politics" for the idea behind this blog. It was written back in 2016, but it's still relevant.

The article touched on how we are being drowned in data so badly that it causes people to not agree as to what is factual anymore.  As well, it talked about how we are producing data at a speed that is so much faster than we are actually able to make sense of it:

One of my favourite quotes from the article:

"It is possible to live in a world of data but no facts. Think of how we employ weather forecasts: We understand that it is not a fact that it will be 75 degrees on Thursday, and that figure will fluctuate all the time. Weather forecasting works in a similar way to sentiment analysis, bringing data from a wide range of sensory devices, and converting this into a constantly evolving narrative about the near future."

So, what does this mean?  If we are able to gather such incredibly accurate data, how are we still unable to make more sense of it?  Why aren't we better at forecasting things like: the day to day weather, election results and the path of the latest hurricane?

The Bottom Line

The world of Big Data was supposed to make things simpler.  With more (and better) data, we were supposed to be able to make better decisions, have society come to a better agreement on things and to improve day to day life.  Instead, have all we done is allow people to better dig in to their point by allowing them to just search out the data that best proves their argument? 

Facebook built one of the first Billion-dollar businesses by using data effectively.  We need to start to use the data that we get now in more effective ways before the real onslaught of data from IoT comes in.....