Could Cloud Computing Lag Make Driverless Cars Unsafe?

September 1, 2017 Off By David
Object Storage
Article Written by Avery Phillips

Any gamer that has given cloud gaming a shot knows that aliens, zombies, and orcs aren’t the biggest enemy they will face – lag is. It can be frustrating to attempt to block an enemy attack only to get decimated by delayed response times. 

But what if this same lag could affect you in real life? What if your car was subject to these same delays? 

In terms of changing the nature of how we interact with the internet, driverless cars are the biggest innovation since the smartphone. And they will be available sooner than you might think. According to major manufacturers like Toyota, Volvo, and Tesla, we should expect them to hit the market before 2020. Is cloud computing, in its current form, able to handle the additional load of thousands upon thousands of new "devices"? Considering the distance that data has to travel to and from driverless cars, this is a real concern. 

Self-driving cars need the reflexes and response times of a professional driver. Safety and even ethical considerations are at play when an autonomous vehicle is placed into hazardous driving conditions. Cloud computing, which commonly has a lag time of one tenth of a second, may not be fast enough. Of course, as with any discussion involving the cloud, data security is of the utmost concern, and the widespread fear of self-driving cars and trucks getting hacked demands answers. 

What is the tech industry’s solution for these obvious problems? How can we ensure that cloud computing lag doesn’t negatively affect the impending first generation of consumer-ready driverless cars? 

The answer lies in a complete change in infrastructure – a change that has been underway and will continue to occur throughout the upcoming years.

A Shift Away From Centralized Server Farms

This concern isn’t a new one, and tech companies have searched for solutions for years. Bare-metal cloud computing has had an impact on the industry for several years, either supplementing or outright replacing "traditional" cloud computing solutions for businesses across the world. However, rigid compliance guidelines and prohibitively high costs (as a long-term choice) have placed this option well out of the reach of most users and small businesses. 

So what is the solution? 

In an article by Wired, Zachary Smith, CEO of New York startup Packet, stated: "It’s a foregone conclusion that giant, centralized server farms that take up 19 city blocks of power are just not going to work everywhere." An obvious solution to reducing lag time is to reduce our reliance on centralized server farms – and bring the servers closer to consumers in more areas. Local server farms could meet the needs of users more quickly and reliably than distant server hubs. By more evenly distributing processing power over a larger area, delays will be minimized. 

Companies like Packet have already begun the groundwork to create the infrastructure needed to support new consumer demands. In stark contrast to most server hubs, this startup has developed a wide array of small datacenters, usually only consisting of a few server racks each. As a result, cloud lag has been reduced from a tenth of a second to around 10 or 15 milliseconds, which could greatly improve the viability of driverless cars, in addition to improved speeds for general communication services like video calls and Wi-Fi calling. 

Fortunately, the decentralization of cloud servers will not have a negative impact on security. Nevertheless, while cybersecurity problems of driverless cars are not exacerbated by a move towards decentralization, they are massive problems. Security must adapt to the internet of things and ensure that consumers can ride safely in these vehicles before they are distributed en masse. 

The cloud computing industry has reached unprecedented levels, nearly exceeding $250 billion. A move to a decentralized system will obviously represent a major shake-up – but to innovators on the cutting edge, seeking to meet consumer needs, this will likely be but a small obstacle. If the industry aims to decentralize major server hubs and provide faster service to users across the globe, autonomous cars will not be impacted by poor response times or compromised security.

##

About the Author

Avery Phillips is a unicorn of a human being who loves all things relating to people and their entrepreneurial spirits. Comment down below or tweet her @a_taylorian.