AI & ETHICS – WHEN AUTONOMOUS VEHICLES MAKE MISTAKES
Auto Tech Review|January 2020
The decision paths within a system with artificial intelligence (AI) are unfathomable. Mistakes are difficult to comprehend, but can lead to serious errors and ethical questions. In the article, Orange Business Services shows why explainable AI is needed for the responsible handling of AI and why blockchain could be a solution.
AI & ETHICS – WHEN AUTONOMOUS VEHICLES MAKE MISTAKES

AUTONOMOUS DRIVING CARS

Driverless cars, 1, train their artificial intelligence (AI) based on neuronal networks in test environments. In this context, it has become evident that the vehicles are prone to errors – errors that were partially difficult to understand and trace and that are extremely problematic because wrong decisions of self-driving cars can ultimately put human life at risk. The AI issues revolving around driverless cars, however, also present additional ethical issues.

How would we deal with the case in which the AI-based software of a driverless car has to decide – in the event of an unavoidable accident – between several evasive manoeuvres, which may incur grave consequences? Should it swerve to the right into a group of pedestrians, or veer to the left onto the opposite lane with heavy traffic? Unfortunately, this scenario is just as difficult for a programmer of AI systems to decide, as it would be for any one of us. What has to be considered from an ethical viewpoint when dealing with AI issues? How can the chances be increased, or how can it be guaranteed that AI-based machines as well as driverless cars make the best decision in the case of doubt? How can potential sources of errors be correctly recognised and prevented?

ARTIFICIAL INTELLIGENCE METHODS

The majority of AI applications for the mass market are extensions of human activity. The applications are very heterogeneous and seemingly inexhaustible: while AI can be used to merely optimise an image, for instance, it can also perform more sophisticated tasks. Take the example of estimating the time required to arrive at a certain place, or AI informing the vehicle owner when maintenance is required. Ultimately, AI can perform very complex tasks such as driving a car.

この記事は Auto Tech Review の January 2020 版に掲載されています。

7 日間の Magzter GOLD 無料トライアルを開始して、何千もの厳選されたプレミアム ストーリー、9,000 以上の雑誌や新聞にアクセスしてください。

この記事は Auto Tech Review の January 2020 版に掲載されています。

7 日間の Magzter GOLD 無料トライアルを開始して、何千もの厳選されたプレミアム ストーリー、9,000 以上の雑誌や新聞にアクセスしてください。

AUTO TECH REVIEWのその他の記事すべて表示
Sound Generation For Enhanced Road Safety
Auto Tech Review

Sound Generation For Enhanced Road Safety

With an ever-increasing number of fully electric and hybrid vehicles entering the market, we are witnessing a change in the acoustic experience in and around the car. The silent nature of the electric powertrain means that other road users miss the familiar acoustic cues that warn them of an approaching vehicle. To remedy this situation, Harman is working on various noise management solutions.

time-read
6 分  |
August 2020
OBJECTIVE ASSESSMENT OF THE SAFETY CONTRIBUTION OF TODAY'S AUTOMOTIVE HEADLAMPS
Auto Tech Review

OBJECTIVE ASSESSMENT OF THE SAFETY CONTRIBUTION OF TODAY'S AUTOMOTIVE HEADLAMPS

Good lighting makes a major contribution to road safety. In 2011, the International Commission on Illumination (CIE) presented an evaluation system for headlamps that can objectively assess the quality of illumination and glare limitation. New light sources like LEDs and lighting functions (glare-free high beam, partial high beam) have been available for several years, so that an extension of the proven CIE method is necessary, which takes into account, among other things, changed switch-on and service lives of new lighting functions and represents a generally understandable evaluation system. An evaluation system presented in the following by the Technische Uni versität Darmstadt, which was developed by an encompassing expert commission, is intended to provide representative data and enable an objective evaluation of headlamp performance.

time-read
8 分  |
August 2020
RAPIDO RE-INNOVATES TO CATER TO POST COVID-19 LAST MILE CONNECTIVITY
Auto Tech Review

RAPIDO RE-INNOVATES TO CATER TO POST COVID-19 LAST MILE CONNECTIVITY

The rapid pace of urbanisation in the country has witnessed cities expanding and subsuming surrounding villages and smaller towns, evolving into Tier II and III cities.

time-read
5 分  |
August 2020
FUNCTIONAL SAFETY STANDARDS TOP PRIORITY FOR FUTURE AUTOMOTIVE DEPLOYMENT
Auto Tech Review

FUNCTIONAL SAFETY STANDARDS TOP PRIORITY FOR FUTURE AUTOMOTIVE DEPLOYMENT

Increasing levels of electronic content in vehicles result in a large amount of software that is required to run the scenes behind the mechanical aspects of a vehicle. Electronic systems in automobiles are part of every section, from the powertrain to comfort and convenience systems as well as all types of safety equipment. This makes electronics and the software running these systems a critical part of the equation for future mobility solutions. In addition, putting standards in place to measure methods of developing, testing and functioning of such electronic systems also becomes highly relevant.

time-read
5 分  |
August 2020
“BYD'S EV STRATEGY TO ENCOMPASS MOST FORMS OF TRANSPORTATION NEEDS IN INDIA”
Auto Tech Review

“BYD'S EV STRATEGY TO ENCOMPASS MOST FORMS OF TRANSPORTATION NEEDS IN INDIA”

The country’s electric vehicle (EV) ecosystem is grappling with a plethora of challenges even as it is striving to wriggle itself out of the ‘nascent stage’ tag. In an exclusive chat with Auto Tech Review, Zhang Jie Ketsu, Executive Director, BYD India shares his perspective on the current EV scenario and how the company intends to progress faster in the Indian automotive industry

time-read
7 分  |
August 2020
SAFETY IN ELECTRIC AND AUTONOMOUS CARS – SIMULATION OF COMPLEX CRASH SCENARIOS
Auto Tech Review

SAFETY IN ELECTRIC AND AUTONOMOUS CARS – SIMULATION OF COMPLEX CRASH SCENARIOS

Electric and autonomous vehicles allow for novel seating arrangements and packaging strategies, presenting new safety challenges. Physical crash testing must be supplemented with virtual simulation to ensure vehicle safety on shorter development cycles. With its Simcenter Madymo MBS software, Siemens provides improved runtimes and the Active Human model, enabling accurate and rapid occupant safety analysis in complex crash scenarios.

time-read
6 分  |
August 2020
KIA SONET – A MIGHTY OFFERING IN THE COMPACT SUV SEGMENT
Auto Tech Review

KIA SONET – A MIGHTY OFFERING IN THE COMPACT SUV SEGMENT

It wouldn't be wrong to say that Kia Motors India has taken the Indian market by storm in less than 12 months of its presence in the market.

time-read
4 分  |
August 2020
NISSAN'S NEW DIRECTION: ARIYA ALL-ELECTRIC
Auto Tech Review

NISSAN'S NEW DIRECTION: ARIYA ALL-ELECTRIC

Marking a key milestone in Nissan’s transformation strategy, Ariya ticks a lot of boxes highlighting the strengths of the global automotive giant.

time-read
5 分  |
August 2020
BETTER SAFETY WITH EMERGENCY BRAKE ASSIST FOR MOTORCYCLES
Auto Tech Review

BETTER SAFETY WITH EMERGENCY BRAKE ASSIST FOR MOTORCYCLES

Rear-end collisions with a slower vehicle suddenly cutting into the lane and intersection accidents with cross-traffic are among the most frequent motorcycle accident types. Continental is now developing an emergency brake assist, which detects an imminent collision with a vehicle ahead or with an obstacle, warns the rider and supports him during braking

time-read
7 分  |
August 2020
ADAS WILL SERVE AS CATALYST FOR IMPROVED VEHICLE SAFETY
Auto Tech Review

ADAS WILL SERVE AS CATALYST FOR IMPROVED VEHICLE SAFETY

The automotive industry across the globe is increasingly focussing on safety.

time-read
5 分  |
August 2020