Driving force

We are inching ever closer to self-driving vehicles becoming a reality. Google has already performed several road tests with prototype models, and the results have been more or less positive. But is the world actually ready for this new phenomenon?

The notion of a car that operates autonomously is attractive to many. On the plus side, they may help to decrease the number of auto accidents through their smart driving technology. In addition, they could prove to be more cost-effective and convenient. However, simply making them “smart” may not be the answer.

In an article published on Wired.com, Andrew NG, chief scientist for Baidu, a company that is currently developing their own autonomous vehicle model, presents an interesting argument. He states that the best method of making self-driving vehicles safer is to avoid having them act like their human-powered equivalents. Rather, the best practice lies in ensuring that they rely on predictive technology to function.

Operating a vehicle is all about the driver’s power of observation. In driving school, we are taught the OEA – Observe, Evaluate, Act – system which helps us to react appropriately to different situations we might encounter while on the road. When an obstacle appears, such as a stalled vehicle or a people running across the street, a human driver can easily rely on their instinct and common sense to identify the issue, and then react in a safe manner. 

If you remove the human driver from the equation, the situation changes significantly. A car that is simply programmed to drive and stop may have difficulty processing occurrences like these. Ng argues that today’s computers lack the programming to deal with these issues, and as such, may do more harm than good.

Despite that, there are some advantages to owning a self-driving vehicle. For example, Ng indicates that they can be designed to have full 360-degree awareness, thus eliminating the problem of blind spots, and they will never face distracted driving issues. Thousands of lives are lost per year due to accidents caused by drunk or distracted drivers, and autonomous vehicles will certainly decrease the number of road accidents.

Predictive data and analytics have been instrumental in helping many companies in the tech space resolve rampant errors or security breaches. The same kind of system can help self-driving vehicles to properly function on our roads. Ng states if the vehicle encounters certain road signs or lane markings on the pavement, it should be able to properly address the situation be recalibrating its speed or remaining in its lane, for example.

So, to answer the question mentioned in the beginning, is the world truly ready for self-driving vehicles? The answer is yes. However, before these mechanical marvels make their official debut, extensive research and development must be conducted to ensure that these vehicles are not only safe, but are programmed to be as predictable as possible.