charliekraus

Hands off the wheel! – Riding to work in an IoT device

Blog Post created by charliekraus on Nov 10, 2015

Anyone who has purchased a car in the past couple of years has no doubt noticed a dramatic increase in driver assistance capabilities: intelligent cruise control, auto-braking, lane departure warning, sensors letting you know where cars are around you, etc. These are all part of a roadmap to fully autonomous driving cars. What this blog will focus on is how dependent auto manufacturers and their customers will be on a high performance, 100% available cloud infrastructure for delivery of updates and communication between cars and cloud services such as mapping, driving conditions, and alerts.

 

So, where are we today? Automobiles are now an IoT (Internet of Things) device as some auto manufacturers are already performing over-the-air software updates to their customer’s cars  that enhance capabilities. Electric car manufacturer Tesla has taken the lead in this area, performing updates that increase battery life, improve performance, and power management that extends driving range. A recent update enabled several driver assist functions that allow their cars to steer themselves and change lanes on highways, but the driver must be touching the steering wheel and be ready to take over control at any time. This point brings up the issue of legislation and regulation of autonomous autos. Current legislation dictates drivers remain in charge of their vehicle and must keep their hands on the wheel in all circumstances. Exemptions have been granted for testing. Google has several self-driving cars on roads near their main campus in full autonomous mode, and in France, Peugeot Citroen had their driverless car journey from Paris to Bordeaux (360 miles) entirely in self-drive mode, automatically maintaining speed to road conditions, and changing lanes to pass slower traffic, without incident. In both cases with a driver aboard and ready to take over control instantly if needed.

 

                                   google car.png                      Peugeot car.png

                             

So, where is this going? What are the motivations for self- driving autos? Safety and convenience are primary drivers. Fatigue and human error are a major factor in collisions. Like now common ABS braking and stability control features, any technology that can assist the driver to avoid accidents seems a useful pursuit. On the convenience front, having the car take over on long highway stretches and in rush hour traffic would mitigate the fatigue factor. This last point is of particular interest to me as an observer and victim of commuting congestion. If autonomous autos prevented slowing down and rubber-necking at even the most minor of road incidents, or improving the pacing of traffic at merging points, it would be a huge benefit.

 

What is the role of Internet connectivity and cloud services in all this? At the most basic level, the software controlling these cars will need to be updated periodically. 100% availability is critical for Cloud-based services such as live traffic alerts for incorporation into navigation system routing calculations, and car-to-car messaging for situational awareness.

 

What are some issues that have arisen in real world testing? Current autonomous test cars are programmed to obey all traffic rules to the letter of the law. Mixing these test cars into traffic where humans don’t always behave by the book is a challenge. A couple of specific situations include four-way stops, where a test car couldn’t get through because its sensors kept waiting for other human drivers to stop completely and let it go. The human drivers were inching forward and making eye contact and getting through, but paralyzing the test car. Another example is smart cruise control, which leaves what is considered safe distance between itself and the car ahead. This space is considered by many human drivers to be enough to squeeze into, and they would, causing the test car to back off to maintain distance1. These and other behavioral situations are expected to be addressed by more aggressive algorithms in the test car’s software.

 

Want to see one in action? Click here and play the imbedded video in the article.

 

I guess it’s time for me to weigh-in with my opinion. Full disclosure – I love to drive! If self-driving cars become a reality and ease commuting traffic flow, I’m all for them for this use case. The rest of my drive time – forget it. Give me twisty 2-lanes along the coast or through scenic countryside, and I want full control. I’m confident I’m not alone in feeling this way. I’m keeping my eye on progress in this area so I’ll know when to nab a non-Internet connected sports car for weekends.

 

No matter how this technology evolves, Limelight has the global network with file and data delivery solutions to handle the most complex workflows to ensure safe driving for human drivers and autonomous autos. Learn more at the Software and Devices Solution site on www.limelight.com

1 http://www.nytimes.com/2015/09/02/technology/personaltech/google-says-its-not-the-driverless-cars-fault-its-other-drivers.html?_r=0

Outcomes