RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2

The Ford Transit Connect Language of Visual Communication for unmanned vehicles

Product
Developers: Ford Motor Company
Date of the premiere of the system: 2017
Last Release Date: 2019/02/06
Branches: Transport

Content

Language of visual communication for unmanned vehicles is the developed Ford for autonomous cars a method of the light notification of people around about intentions. In fact, it is attempt to create universal standard visual language for "UAVs" which would be clear to everyone.

2019: Carrying out tests of "light language" UAVs

On February 6, 2019 the Ford company announced carrying out tests of "light language" UAVs within development of the interface of communication which will help autonomous cars to join the general road flow smoothly. Approach of Ford consists in use of light signals using which the UAV announces what it does at the moment and what will be its following action.

Ford found optimal color for "light language" of UAVs

As reported were carried out to Ford, tests supplementing a series of tests in the USA together with the Technology university of Chemnitz in Germany. This time researchers expanded the program of tests, having turned on efficiency evaluation of signals of two complementaries in addition to white, purple and turquoise. Besides, the light panel was placed on a roof, but not in an upper part of a windshield as in the USA, and when modeling road situations the distance between participants of an experiment and the UAV was increased and made up to 500 m.

Tests showed that 60% from 173 people polled after interaction with "UAV" Ford Transit Connect decided that before them there was an autonomous vehicle moving without the assistance of the driver. Reactions of these people and also 1600 more, become involuntary participants of an experiment, helped to reveal optimal color for communication of the UAV with people around. It was it turquoise: it was more noticeable, than white, and at the same time it it was impossible to confuse with red special signals of emergency services as it happened to magenta signals.

Tests also showed the high level of acceptance of signals and trust to them. Thus, the base which will allow researchers to develop and perfect communication using light signals is put, consider in Ford.

As much as possible to bring closer test conditions to real, in Ford created "suit chair" which was set on a driver's seat in the Transit Connect van. The suit with the driver looks for people around as an empty driver's seat inside thanks to what observers could collect more reliable information about reaction of people to the light panel fixed on a roof which using signals of white, violet and turquoise color notified that the van begins the movement, moves or gives way.

File:Aquote1.png
"The visual contact (with the driver) is important, but our research showed that first of all participants of traffic watch what does the car. Now we will try to understand how it is possible to make light signals the clearest and intuitive for all".

Matthias Beggiato from department of psychology at the university with which Ford worked on the InMotion project
File:Aquote2.png

According to Ford, for financing of the project the grant in 1 million euros with the Ministry of Transport and Digital Infrastructure of Germany within a study program on automation and communications on the motor transport was selected.

Drivers of "UAVs" were trained to safe driving and monitored a situation on the road through a false head restraint. The light signal was given by them manually using the special lever. The assistant disappearing on a back seat also monitored the road and provided the driver with water.

With the purpose to bring to the market the USA by 2021 the autonomous car which is initially designed for pilotless use Ford works on increase in trust to UAVs. According to the developer, the industrial standard of signals using which the autonomous vehicle notifies other participants of traffic on the intentions is for this purpose extremely necessary. The company cooperates with several industry organizations, including with International Organization for Standardization and Society of engineers of automotive industry, and urges other automobile and technology companies to help with creation of this standard.

Fords in partnership with the company Argo AI which is engaged in development of autonomous cars carry out tests of UAVs in the capital of the USA Washington and also in Detroit, Pittsburgh and Miami. In China the company cooperates with local searcher Baidu within the Apollo program which provides tests of UAVs on the selected roads in Beijing and other Chinese cities which will begin in 2019.

2017: Studying of reaction of people on highways on the autonomous car

On September 18, 2017 Ford Motor Company announced carrying out together with the Virginia institute of the equipment and transport of a research of real reaction of people on highways on the autonomous car signaling about the intentions.

Research purpose

Ford tries to study all range of effects of emergence of autonomous cars on public roads in the future to ensure safety of joint coexistence of the "UAVs", pedestrians and machines managed by people. The research, joint with institute, which purpose — search of the most effective remedies of communication between the autonomous car and other participants of traffic became a part of these efforts also.

Arguments for benefit of light signals

Researchers considered the possibility of an output of text messages to the display, but it would demand knowledge of a certain language all people around. Characters were rejected as use the low level of recognition among users.

As a result scientists addressed light signals. The light signals announcing braking or intention to make turn, are already standardized and easily understood. Therefore light indication was considered the most effective remedy for creation of the protocol of visual communication for autonomous cars. Specialists of Ford developed light signals which will announce people around that the car moves in a standalone mode, gives way or it is ready to begin the movement after a stop.

Preparation for an experiment

For this purpose over a windshield of the Ford Transit Connect car fixed a light panel. Six video cameras of high resolution provided circular shooting of reaction of people around. To imitate the movement in completely standalone mode, researchers also developed the innovation chair suit hiding the person driving. The suit with the person looks as a normal automobile chair inside and creates illusion of completely autonomous car without driver. This simulation was irreplaceable for assessment of reaction of people around and their behavior at a meeting with "UAV" in the real urban environment.

Testing of light signals

After that the team of researchers started an experiment on use of three different types of light signals for the notification of people around about "intentions" of artificial intelligence of the car:

  • I give way: two white rectangles moving here and there. Show that the car is ready to give way up to a complete stop;
  • The movement in a standalone mode: the continuous unblinking this world meaning that the car moves in completely standalone mode;
  • Start of motion: quickly blinking this world meaning that the car starts after a complete stop.

Ford Transit Connect with the driver in a chair suit imitated autonomous movement along public roads in a densely populated northern part of the State of Virginia for all August. During the experiment researchers recorded video and kept diaries of observations of reaction of people. The car passed about 1800 miles in the urban environment, over 150 hours were during this time shot by videos on which interaction of the car with pedestrians, cyclists and drivers of other vehicles is imprinted. Light signals were used more than 1650 times, including at intersections, parkings, in garages and on outcomes at the airport.

It is expected that collected data will help researchers of Ford and scientists to understand how the behavior of participants of the movement in response to signals of the autonomous car changes.

Robotics