There are specifically two trends that are happening regardless of having an autonomous vehicle today. They are basically; Driving assist and over the air updates
Driving assist: It is being regulated . Regulatory bodies are incentivizing car manufacturers to build cars with driving assist sensors as well as collusion avoidance sensors in manufactured cars. In fact by 2020 every new car will be equipped with a front facing camera to satisfy regulatory bodies in order to get a five-star rating[8]. For example in 2015 Mobileye company shipped 4.5 million censored cars with front facing camera and powerful chips to process the visual data coming from it. This meet the regulatory bodies requirements [8].
Over the air updates: This actually started
…show more content…
These two big challenges or problems are ; Environmental Model and Path Planning.
Environmental Model: This model deals with the ability of the system to tell ultimately for every pixel which appears in the image what category it comes from. For example, if a certain pixel is coming from a barrier, coming from the side of a car or if the pixel is coming from the pedestrian and so on. This system should be able to tell where any of the pixel is coming from in order for the driver to be able to build this environmental model so as to be able to tell the free space where it can drive and where it is not allowed to drive [6].
Path Planning: This is a fuse of all information available from the image and not only lane marks. It definitely expects the system to be able to tell a driving path direction on where exactly to drive through.
These two models that are exactly the major problems facing autonomous driving today. Measures are being made to resolve these problem and adequate solutions would soon be found soon in the next few years. The tool that is being used in resolving these problems is called "deep learning"
…show more content…
summarizing the development of these vehicles, there is already in 2015-2017 high way autopilot and testing of these automated vehicles. They are not fully autonomous vehicles. They are not so much to be seen on our roads today but are still being tested for their accuracy.
2018- 2020 will witness some percentage of highly autonomous driving that have about 360 degree awareness. But these vehicles will have actuations only on highways and can only be activated only on high ways with a grace period of about 30 seconds to take over control by the human driver. These vehicles will be designed to act on its own, and alert the driver to take control when necessary [8]. It is mainly designed to be safe during high way driving and its coming on our roads between 2018 and 2020.
The level 4 and level 5 vehicles which are regarded as the fully autonomous driving will be coming to our roads in 2021. These are fleets of cars that might not be sold to the public. They are generally designed to operate in geo fenced areas with the purpose of building a new business model. This technology will operate in the urban cities under shared mobility