Autonomous Driving Development Using Cloud Computing

in Project HOPE5 years ago

We are in 21st Century and days are not far away when there will be no need for a driver in our car. Driverless cars are not a new thing for us now.

For this automation Cloud technology & Data is the lifeline of Autonomous Driving development, Management, analytics, and reuse has a major impact on developmental costs & timelines as well.

There is a drastic change in the automotive industry and it is going through a ‘Tech-Tonic’ revolution with a major focus on Autonomous Driving (AD) development in Cloud.

edge-computing-in-an-autonomous-car.jpg
Image Source

An Autonomous vehicle (AV) calculates the input data by receiving from different sensors to take the driving decision. In normal conditions, a typical autonomous vehicle produces upwards of 40 TB of data in 1 day and this data is ‘Gold’ for autonomous driving development.

The data generated whilst testing AVs is required to be stored & reused to train & refine the Autonomous Driving software & different AI, Deep Learning, Machine Learning software equipment.

automotive-smart-car.jpg
Image Source

There is a dedicated focus on the administration of data and computation of dynamic-distributed data, the reason is that it can have a critical impact & costs can rise exponentially if data & computation are ineffectively managed with raising complexity in the vehicle development cycle.

It is essential and needs to ensure that all the data generated should as useful or not useful after its classification so that it can be reused for the development process to cost and to save time as well.

Moreover, additional data or we can say Petabytes is generated during the downstream development systems & process for example when a person executes validation processes like Processor in Loop (PIL), Model in Loop (MIL), Hardware in Loop (HIL), and Software in Loop (SIL).

image-asset.jpeg
Image Source

Every version of this validation test must be stored & should be accessible to the engineers so that the algorithms can be evaluated & it can be fine-tuned.

So, these activities should be implemented at ‘Hyper-Scale’ for example, an autonomous vehicle should be tested in the virtual world or in simulation against millions of scenarios & if a virtual scenario is two minutes long it will take thousand days to execute, but at the ‘Hyper Scale’ one can execute this in just ten days.

This automotive industry is in the starting era of implementing this Cloud technology and techniques and there is a lot that will be done in the future in terms of innovations and in terms of applications.

@printskill


CC to: @project.hope @crypto.piotr @juanmolina @lanzjoseg @fucho80 @jadams2k18 @pedrobrito2004 @content-creators

Plag.png

Sort:  


image.png

Your post has been upvoted by @zero-to-infinity. We are supporting all the STEM Content Publish in Steemit.

For more,you can visit this community

JOIN WITH US ON DISCORD SERVER:

Support us by delegating STEEM POWER.
20 SP50 SP100 SP250 SP500 SP

Follow @zero-to-infinity & @steemitblog for latest updates

Thank you so much @zero-to-infinity for your support.

Coin Marketplace

STEEM 0.04
TRX 0.33
JST 0.081
BTC 62845.62
ETH 1666.27
USDT 1.00
SBD 0.42