While we agree that there is a space wide open for autonomous cars to flourish but we also believe that it’s going to be a roller coaster ride for the hardware suppliers. Let’s discuss them one by one below.

 

1. The slow pace of EV industry in US:

 

An autonomous car has to be Electric or else there will be added burden to programme it to be able to fuel itself and change gears. The US electric vehicle market is slowing down as American car buyers have been slower to purchase Electric vehicles. It’s estimated that EV sales in China in 2017 stormed past 400,000 units, while Europe sold about 135,000 EVs. Meanwhile, US buyers purchased only about 116,000 electric cars. Considering the fact that US is the hotbed of autonomous car development, the numbers are plain abysmal especially when the more than ~17 million new cars found new homes in 2017.

 

 

“Americans find Electric and Plug-in hybrids vehicles prohibitively expensive”

 

The biggest obstacle in US for electric vehicles wide adoption is their failure to address an actual problem from the driver’s point of view. Electric vehicles have less range as compared to a gas engine, have a higher initial cost and slow charging time. Internal-combustion engine powered cars offer a more-viable transportation option. It’s a common understanding that for a new technology to replace older technology, it needs to address an existing inadequacy, solve a problem or at least offer similar service at somewhat lower cost. Electric vehicles fail on all these measures we listed earlier, thus the EV market has not grown to its potential in the US.

 

2.  Cybersecurity:

 

In today’s interconnected world, hacking is an omnipresent danger. Hackers can break into systems to steal personal. Financial information, bring down targeted websites, access sensitive government websites and what not. Now imagine, if instead of targeting desktop computers, websites or even mobile phones that such hackers could target autonomous cars which are fully run by different sensors and cameras are ultimately controlled by software, and a software is always vulnerable to hacking.

 

The main idea behind autonomous cars is to reduce fatalities caused due to human errors but if hackers start to hack them, there will be furthermore accidents where there will be no meaning for the autonomous cars, that is a lot of vulnerability and a lot of risks to contend with and for this reason cyber-security is a core challenge for self-driving cars and its components.

 

An autonomous car has 60 to 140 electronic control units’ basically tiny computers each having multi-million lines of code. Every 1,000 lines in the code contain as many as 15 bugs which could be a potential target for hackers. The idea of extracting more efficiency and responsiveness from a device by virtue Internet of Things (IoT) is resulting in a growing list of devices being connected to the internet and self-driving cars will only add to it. This connectivity to the internet greatly increases the potential for the car itself to become the target of cyber-attack.

 

3. Uncertain future of autonomous cars:

 

Let’s face it, Autonomous vehicles are the buzzword right now, but it’s still unclear whether they will be in favor or against the future of drivers, automakers, tire-1 manufacturers and tier-1 retailers. As discussed earlier, Autonomous cars will rely on a lot of technology, both onboard and external, such as GPS and radar, LIDARs and cameras to guide them. The companies are still working on AI-artificial intelligence and deep learning for improving this technology to be as competent as a human brain but reaching there will take time.

 

There are many differing opinions on when will autonomous vehicles be deployed. This is due to the complexity of the engineering solutions that must be integrated and delivered on an automotive platform. Many technical challenges remain to be solved: algorithmic, technological, and societal. In all these topical areas, a common and fundamental feature is associated with the control of uncertainties in the sensors, algorithms, and people’s use and a tendency for use of these technologies.

 

Autonomous cars are equipped with a large variety of sensors, including GPS, the variety of cameras, LIDAR, proximity sensors and others. The data coming from sensors is not always accurate and that must be accounted for in subsequent computations that fuse it and build higher-level representations of scenes and situations.

 

Simple measures of inaccuracies are the sensor resolution and the sensor noise model. The complexities don’t end here as the fusion of multiple sensors in which an embedded reasoning system is required to understand the inherent quality of the sensors along with the change in environments.The new generation sensors are small systems with highly complex algorithms, which further reduce the ability to separately qualify the sensor operation under changing environmental conditions due to the add-on computing functionalities.

 

Also read :  Self-Driving Cars: Liability and Insurance

 

4.  High Risks Associated with Autonomous Components:

 

In self-driving cars, software relies on various disciplines, such as machine learning, computer vision, and parallel computing. It is a complex process to replicate a human brain i.e. make a decision, and what is even more difficult is to test those against all possible real-world scenarios.

 

Self-driving car Engineers need to actively decide on the kind of data they want to use, their trustworthiness is and the biggest of all, how to balance the myriad sources of information in their algorithms. The problem is usually referred to as sensor fusion. This problem is vital in the case of connected vehicles as in V-2-X connectivity, data will come not only from the sensors of the car but also from other vehicles, street infrastructure, etc. In this case, other factors should be taken into account since it is not possible to have a perfect knowledge of the devices that are used to sense information and about their status.

 

 

There is a question that what if the components fail or mislead the cars?

 

For instance, on not so perfect cloudy day weather conditions, the navigation might report a clear street ahead, the radar might be reporting a clear street, but the visual camera could show an obstacle in the path. How will this “equation” be solved and what will be the result? One wrong decision might lead to an accident, and the very purpose of creating an autonomous vehicle society will be defeated. Although, Car manufacturers and numerous start-ups active in this space are constantly improving and testing the recognition capabilities of their systems. But.it is a multi-factor optimization task, which requires them to find an optimal solution under consideration of costs, quality, and potential risks.

 

The above information has been sourced from our report titled “GLOBAL SELF-DRIVING CAR HARDWARE MARKET . Download free sample to know more