There are still a few years remaining before fully driverless cars will hit the road in significant numbers. Speed of adoption of this technology will depend on how fast the industry and governments can break down the barriers to adoption.
The financial barrier
One of the main barriers to adoption is cost. In 2010, Google’s self-driving technology cost around $150,000, of which $70,000 was just for LIDAR remote sensing technology. Obviously such a price tag is dissuasive, but mass production and R&D can do miracles; for instance Quanergy Systems announced last month that it will be introducing in 2016 a solid state LIDAR for less than $1,000.
Many car manufacturers, together with their OEMs, are looking at ways to consolidate and simplify the hardware. Many startups are also looking at a piece of the action, focusing on miniaturisation, sensor fusion and integration of controllers as potential solutions. Better optical systems that could replace LIDAR and radar technologies are also envisioned. Computational processing, which is still another large component of the overall price, will experience the usual exponential cost reduction.
Whatever the technological solutions that will prevail, one thing is certain – cost will go down. How quickly is still a question mark, but according to a recent studyfrom HIS Automotive, price for the self-driving technology will add between $7,000 and $10,000 to a car’s price in 2025, a figure that will drop to around $5,000 in 2030 and about $3,000 in 2035, the year when the report says most self-driving vehicles will be operated independently from a human occupant’s control.
Some other reports suggest that by 2030, the price of the additional technologies will fall under $1,000, at which point the autonomous option will cost probably less than the annual savings in insurance.
The legal barrier
As costs will come down inexorably, the next most important barrier is legal. In my opinion, and though these legal issues are real, they will be tackled in time to allow the advent of unmanned cars sooner than later.
The fact is that some US states, such as Nevada, Florida, California and Michigan, and countries such as Japan, the UK and France, have already created a legal framework to allow self-driving cars. Unfortunately, this framework is there more as a permission to test vehicles rather than creating the real legal foundation for driverless technology to share the roads with current cars.
These laws acknowledge that vehicle “operators” don’t need to pay attention but must avoid disturbing activities such as sendingtext messages as well as requiring a person behind the wheel at all time. As economical interests will prevail, more and more countries and states will legalise, in some form, self-driving cars. The question is: when will they tackle the real fundamental change originating from having no human being behind the steering wheel and the impossibility to incarcerate a robot car?
New laws will have to tackle this reality by defining the shared and individual liabilities between the future players: car manufacturers, onboard signaling system providers, car-to way-side communication providers (including V2V, V2C and V2I), control centre hosting the network information, road infrastructure providers and the car operator or owner.
The barrier of agreeing on safety standards
As long as driverless cars will be much safer than conventional cars, governments will find the way to redefine existing laws.
To be safer, driverless cars will need to adopt in a form or another the four safety principles we’ve described previously: block interlocking, forbidding two cars to be in the same zone at the same time; block signaling indicating to the next vehicles that no vehicle is in that road section; road and carriage integrity ensuring that no vehicles as departed from the driven car and that the road is in operational conditions; interoperability, ensuring that physical and conceptual integration is possible.
All stakeholders will need to agree on how these safety principles will need to apply to all system elements, a daunting challenge for an industry well known for poor cooperation between major players.
Furthermore, when these conditions won’t be met, cars will need to adopt safety procedures or rely on fail-safe modes and redundancy features to take over, allowing vehicles to continue to run or stop safely. Authorities and homologation bodies will need to help define what these rules will be under all potential conditions. As lives are at stake, safety levels will need to be defined according to the expected automation level, with their testing procedures and integrated within standards.
The barrier of trust
For experienced drivers, putting their life into the hands of their car’s onboard computer won’t be easy. As passengers, they will need to trust their car as well as neighbouring vehicles in a few different ways.
From a telecom perspective, human beings will need to trust that the received signal comes from the car directly in front or on its side, and isn’t trying to hack into the car’s system or being sent by malicious people. Additionally they will need to trust that other cars are accurately reporting their state and not sending wrong messages.
From an onboard perspective, passengers will need to trust that the safety concepts are met, the algorithms are able to consider all potential situations accurately and that their car will react in a safe way, in case of failure. These are a few of the worries passengers will have, but as long as driverless cars can show that they are safer than humans, trust will be granted.
Government initiatives to eliminate barriers
Most countries with an automotive industry are pushing initiatives to position well their local champions on such a fundamental change of paradigm. The EU Government is tackling the barriers to adoption within its 2020 Horizon plan.
Through its mobility for growth 2016 and 2017 work program, it will be allocating funds to safety and intelligent transport, including a 25m Euro large-scale demonstration of cooperative ITS (work package: MG-6.2-2016).
Furthermore, it is also earmarking in its call “Automated Road Transport” 64M (in 2016)and 50M euro (in 2017) for projects helping define how to break-down these barriers:
- ART-01-2017: ICT infrastructure to enable the transition towards road transport automation
- ART-02-2016: Automation pilots for passenger cars
- ART-03-2017: Multi-Brand platooning in real traffic conditions
- ART-04-2016: Safety and end-user acceptance aspects of road automation in the transition period
- ART-05-2016: Road infrastructure to support the transition to automation and the coexistence of conventional and automated vehicles on the same network
- ART-06-2016: Coordination of activities in support of road automation
- ART-07-2017: Full-scale demonstration of urban road transport automation
For the SMEs interested in receiving funds by participating in these programs, you can find out more here. However, you need to hurry up as the deadline for submitting your project is January 2016 for many work packages.
Source:ConnectedCarTech
No comments:
Post a Comment