Category: Autonomous Driving

A Street Car You Desire

wiki-horse-street-carThere is something called “desire paths”, that I learned about while watching an excellent TED talk by Mr. Tom Hulme . Desire paths are the short-cuts or “paths of least resistance” that one recognizes while interacting with a structured model. Sidewalk landscaping around buildings is a good example. Have you ever found that the walkways to buildings are too circuitous, taking you on unnecessary journeys through gardens and parking lots? They abdicate the “line” rule, the shortest distance between 2 points. Impatient people like me may cut through the lawn and  wear down the grass until a more direct dirt path emerges. If the architected field of dreams does not correctly anticipate what we need,  users may not come. 

Recognition of desire paths may improve implementation of any technology. When the elevator was first developed, there was nervousness about the cable breaking and the risk of a free fall for all.  Innovators, like Elisha Otis in 1852 pioneered solutions. Now, the pleasantries and culture of attendant operated elevators are forgotten and automation is taken for granted. The designers of the autonomous car also believe that we will learn to accept and safely use their technology too.

So what are desire paths that may enable quicker implementation of the autonomous car? One way is to put smart cars on smart roads. This means switching on autonomy when the enabled vehicle drives on a road that can interact with it because of technology integrated in the roadway, signs, and lighting. Highway lanes could be designated for autonomous vehicles just as there are lanes for vehicles with more than one passenger.   When the vehicle leaves this special lane, it would revert to manual control. 

Another model for autonomous driving is vehicle platooning. Here, cars or trucks could be
switched into autonomous mode when they join a string of similar vehicles. This sequence of vehicles resembles the attached cars of a train. The first vehicle in the platoon is manually driven. In turn, it chauffeurs the vehicles behind it. This model is being investigated by the Safe Road Trains for the Environment (SARTRE) project in Europe. Advantages include fuel efficiency, decreased wind drag, and autonomy for the chauffeured vehicles. 

Perhaps someday, I will own a car with a single red brake button and no steering wheel or floor pedals. And I may find special desire paths for driving such a car. Until that day comes, there are other ways autonomous vehicles will become mainstream soon – at least that’s what I desire.

Formula One race car with light effect. Race car with no brand name is designed and modelled by myself

References: The concept of “autopilot” lanes was described in an article by T Melba Kurman, Triple Helix Innovation and Hod Lipson, Cornell University in December 2013 called: Where Are the Autopilot Lanes for Driverless Cars? (Op-Ed) 

Future Car from iStock photo. Formula One race car with light effect.

Horse drawn street car: “Rapid transit in 1877″ – First horsecar run in Manchester, New Hampshire”. Published 1908 by the Hugh C. Leighton Company, Portland, Maine. Image was downloaded from Wikimedia Commons.

Blood Flow As a Model for Autonomous Transportation

Nature has a way of recreating biologic events in surprising ways, such as the coin-like stacking of red cells and the linear linking of autonomous cars moving through arterial highways. Red cells are round and similarly shaped and can connect to one another by
protein links in a stack of discs, called “rouleaux”.  Greys RBCs 2

Autonomous vehicles might similarly connect in platoons to increase roadway capacity and transportation efficiency. Good-bye the 2 second safety spacing rule between non-autonomous cars!  Instead, autonomous cars, linked bumper to bumper, will move as one, like stacked blood cells. When the light turns green, the line of connected cars will move through the intersection together. Just as red cells may collect in rouleaux or disconnect to move individually, the autonomous car can join or leave a sequence of platoons according to destination.

The study of blood can lend other lessons to future transportation planning, such as laminar and turbulent flow, cellular diversity, rheology, and pathology. Understanding these factors may spur novel approaches to vehicle and roadway design and anticipate imperfection and disease. Driving through highway construction traffic, I learned of Ford’s plan to mass produce autonomous cars in five years. Sounds like a bloody good idea to me!

Reference: The above image is from Henry Gray’s Anatomy of the Human Body (1918) via Wikipedia Commons. Panel “a” shows red blood cells en face. The cell has a discoid or bi-concave form that maximizes its surface area, which may be more important for laminar flow than diffusion of oxygen.  Laminar flow is orderly flow in parallel layers, rather than the disorganized motion of particles moving in different directions and velocities seen in turbulent flow. If most cars on the road adopted the same size and shape, speed, and direction, more efficient laminar-like flow would be expected. Autonomous vehicles in rouleaux formation might take the turbulent steam out of road rage and other erratic driving styles. Panel “b” shows red cells stacked in rouleaux formation, which can be visualized on ultrasound or echo images of the heart and blood vessels as smoke.