An automotive revolution is upon us, one that will give new meaning to the term “back seat driver.” A recent report by Boston Global Consulting predicts that the autonomous or self-driving car market will hit $42 billion USD and create 100,000 new jobs by 2025. Automotive and tech companies are harnessing the power of mobile networks to change cars as we know them. And the future, it seems, is fast approaching.

What is a self-driving car?

Frequently referred to as ‘autonomous vehicles’, self-driving cars have the ability to assess an environment and navigate it without human input. Leveraging breakthrough IOT (Internet of Things) and AI (artificial intelligence) technology, these cars can ‘make their own decisions’ and are set to transform both the personal vehicle and car share markets.

In the US, where much autonomous car testing is taking place, the Society of Automotive Engineers (SAE) has developed a level rating for self-driving vehicles. Starting at zero, the spectrum then ranges from ‘one’ (driver assisting automation, like cruise control) to ‘five’ (a car that handles all driving tasks, you just get in and tell it where to go).

What’s on the road?

In the personal automotive sector, we’ve only hit ‘level two’ on-road. A level two car, such as Tesla’s ‘autopilot’ system, can activate cruise control, stay in its lane, and also slow down to avoid other vehicles. But major automotive companies are moving fast to build cars and acquire businesses that will be vital to the autonomous landscape. Earlier this year, GM invested $500 million in Uber-rival Lyft and acquired car-sharing platform Maven. Ford, a serious player in the autonomous category, invested in sensor and mapping-company Velodyne, algorithm providers SAIPS, AI company Nirenberg Neuroscience and 3D mappers Civil Maps. Daimler, which owns Mercedes-Benz, threw money into route planning app Moovel, and car share company Car2Go.

Tech companies are also in the race. Google’s Self Driving Car Project, or Waymo, is testing a combination of maps, sensors and software to create vehicles that can autonomously plot geographical information, navigate around other objects (including vehicles, cyclists and pedestrians), and predict what these objects might do next. Based on the info, the software also makes decisions surround speed and direction.

And then, of course, there is Tesla. In early 2016 the company rolled out its 7.1 “autopilot” software for Model S and Model X cars, which offered safety features like automatic lane change and collision avoidance. Autopilot 8.0 includes more responsive and traffic-aware cruise control and auto steer. One of Tesla’s significant advantages over other automotive and tech companies is its ability to harness mobile networks and send autopilot updates over the air. Tesla drivers can update cars at the push of a button, not a trip to the mechanic.

The downside of Tesla and other autonomous automotive? The cost.

An autopilot-enabled Tesla hits upwards of US$77,000, which is why ride-sharing services are attracting so much attention in this space. Unsurprisingly, Uber is a big player with Robo-Cars already providing rides in US cities like Pittsburgh and Pennsylvania.

The self-driving car revolution relies on the establishing of a network that can support communication between cars. As Sebastian Zimmerman of BMW says, “you need ultra-reliable networks, low-latency and they must work everywhere.” 5G networks, the next evolution of 4G, provides the necessary bassline for fast communication between self-driving cars (as well as other IoT applications, like smart homes). How crucial is 5G? It’s 50 times faster than 4G, meaning if cars detect an accident they’ll be able to react and notify 50 cars around them in the time a 4G-equipped car could notify one.

Pros and cons

Autonomous vehicles might mean fewer drivers but they also could mean more jobs. Online educator Udacity has teamed up with Mercedes-Benz, Nvidia, Chinese ridesharers Didi, and Uber’s recently acquired truck company Otto to launch a 27-week course on autonomous engineering. Instructing in basic coding and deep learning, the course takes students through tasks needed to make a self-driving car, including path planning, lane line detection and non-GPS dependent location mapping.

But as the industry accelerates ahead, so do conversations around its “algorithmic morality.” If a vehicle is able to make decisions, what happens when it’s faced with an unavoidable accident? In 2014, over 1200 people died in Australian road accidents; in America the number was over 30,000. Machines don’t drink and drive, or text on the road, which makes them potentially safer drivers than humans, but in the case of an accident, could and should a vehicle choose whom to prioritise? These questions are becoming vital as we gear up for level-five driverless vehicles. For now, however, the majority of us can sit back and enjoy the ride.

4 Minute Read

Read More

Aimee Chanthadavong

Content Producer

Aimee Chanthadavong,
Content Producer

As Content Producer of RedWire, Aimee is a passionate storyteller about people, technology, and anything else that requires her to use a bit of journalistic detective work.