Out of loop? Out of question!

What a minor driving mishap can warn us about our fetish for self driving cars!

On a recent road trip in a rental car, stuck in stop-and-go traffic, to ease the cramps in my overworked right leg, I tried something utterly stupid: I tried left foot braking. What followed was enough to clearly drive home the argument that this technique is recommended only to the most adept of racing drivers. Instead of gently tapping on the brake pedal, I slammed my left foot hard on it, bringing the car to a grinding halt. How I managed to not be rear ended stands to be a miracle indeed. To avoid embarrassing myself, I'll not get into the panic that ensued.

The day's lesson was clear enough - my left leg just didn't have the motor control or reception of tactile feedback to safely engage the brake pedal. This is strictly given how "out of loop" my left foot is from the driving process, both traditionally and in that particular instant. Mind you, I drive stick regularly, where I'm constantly using my left foot in the driving process to control the clutch (the rental car was an automatic). In fact, the reason why drivers of vehicles with automatic transmissions are suggested to use only one leg to control acceleration and braking is this: to enable a constant channel of feedback about the vehicle's actions while reducing the cognitive needs associated with using two feet.  In other words, to effectively stay "in the loop". 

Self driving cars are here! (Not really)
Cut to the world of today, vehicles with driving assistance systems and semi-autonomous features are increasingly more prevalent. Everyone from ride hailing services to automotive manufacturers and software companies are bullish on being able to bring fully autonomous vehicles to market and thereby take human driving out of the equation. Much is being said about the transitionary phase - from where a driver controls all aspects of the vehicle to where drivers can take the backseat and flip through a magazine as the car ferries one to a destination.

Here are some practical facets of this paradigm: 

• If claims are to be believed, manufacturers currently have the basic technology (imaging technology, obstacle sensing, vehicle-vehicle and vehicle-infra communications) that is capable of automating some or most aspects of the driving task, in controlled situations.
• This is strictly constrained by the infrastructure - the condition of road signage and markings, and the ability or lack thereof of the vehicle communicating with the surrounding vehicles and infrastructure to predict oncoming scenarios.
• Verdict is still out on the performance of this technology in extreme weather conditions where visibility and obstacle sensing is hindered. 
• Above all, the technology is still in a situation where it needs to hand the driver control of the vehicle atleast 'some of the time'. 

This list can ascertain the strong probability of driver-vehicle control handoffs, situations where a semi-autonomous vehicle needs the driver to take over driving control, in the times to come. However, going by marketing, journalistic, and PR buzzwordry, one might assume its safe to just let go off the wheel and pedals in a car, in the nearer future. When that moment does come, I reckon a safe driver-vehicle handoff isn't as easy as the current discourse would have one believe. It's an incredibly complex problem that needs a nuanced solution that is carefully researched and deployed.

Going by my own anecdote, bringing merely one "out of loop" body part into the driving loop has led to, let's just say, uneasy consequences. That just goes to say, until vehicles get completely reliable full-autonomy - taking drivers out of the driving loop needs to be out of question!