Column: Automotive lighting and self-driving vehicle safety
After more than a century the motor car is undergoing a technological transformation such as we’ve never seen before.
Just about every aspect of the design and functionality of the vehicle is being redefined – the powertrain is being electrified, computing power is replacing horsepower, and safety is moving from active and passive systems to intelligent integrated designs. Among which are a complete rethink of interior and exterior lighting.
And as vehicles become increasingly automated as they travel the path to full autonomy, the role of lighting is shifting from cabin and road illumination and basic indication through turn signals and brake lights, to driver engagement and even communication with the vehicle’s surroundings.
LEDs enhance safety in ZF’s smart steering wheel
While Level 3 automated vehicles are undoubtedly safer than human-piloted cars in most situations, they nevertheless require the driver to be engaged and prepared to take back control at all times. This engagement and handover is considered by many experts to be the weak link in L3 and L4 automation.
According to Jürgen Krebs, vice president of engineering for steering wheel systems and driver airbags at Tier One supplier ZF: “As ZF pursues its goal of “Vision Zero”, a critical enabler will be vehicle and driver interface.
“As new automated functions become more commonplace, advanced technologies employed in the steering wheel are important and can help improve driver safety and awareness of the current vehicle control mode.”
Thus the company has produced a concept steering wheel that uses an LED light strip embedded in the rim to communicate with the driver by way of changing color:
- Blue lights – autonomous mode is engaged
- White lights – the vehicle is in manual driving mode
- Yellow lights – left or right-hand turn signals
- Red lights – alert the driver to potential hazards
In order to determine the level of driver engagement the system uses 10 capacitive sensors in the outside rim that detect where the driver’s hands are placed on the wheel, with an additional sensor on the inside of the rim used to detect whether an appropriate grip is being employed.
Critical for safe L3 driving, the system enables accurate hands on or off detection while communicating to the driver which driving mode is operational.
While Level 3 and 4 automation pose many challenges regarding driver-machine interface, self-driving cars come with an entirely different set of concerns. With no controls, and in some cases no occupants, ‘Robo-cars’ could be perceived as unpredictable and even threatening by other road users.
Lighting is the new sign-language for self-driving cars
While traffic signals, signs, and road markings provide explicit guidelines for those using roadways, some decisions, such as determining who will go first, are mostly made through unspoken negotiations between road users.
Sucha et al (2017) found that a pedestrian’s decision to cross a road as well as their feeling of safety are affected by various signals given by the driver; such as eye contact, waving a hand, posture, and flashing lights.
The study showed that 84 percent of pedestrians sought eye contact with drivers. Schmidt and Färber (2009) found that pedestrians who want to cross the street look at the approaching driver to get acknowledgment from the driver – for example if the driver returns the eye contact, pedestrians assume that they have been seen and that they have achieved mutual understanding.
However, in the case of driverless self-driving vehicles this no longer applies, and needs to be replaced with another form of machine-to-human communication and connection.
An alternative form of interface, proposed by Ford, could take the form of a light display that would signal the driverless vehicle’s status and intent to other road users.
To validate the concept Ford used a Transit Connect, disguised to look like a self-driving car, to conduct real-world trials. The vehicle was equipped with a light bar above the windshield, positioned where a pedestrian or cyclist would normally look to make eye contact with the vehicle’s driver.
To signal status and intent the light bar was pre-programed for the following scenarios:
- Yielding – Two white lights moving side to side to indicate vehicle is about to come to a full stop
- Active driving mode – Solid white light to indicate vehicle intends to proceed on its current course (although can respond appropriately to objects and other road users in the course of its travel)
- Start-to-go – Rapidly blinking white light to indicate the vehicle is beginning to accelerate from a stop
Interestingly the trial showed that with no prior explanation of what the different signals meant, it took about two exposures for participants to learn what a single signal meant and between five and 10 exposures to understand the meaning of all three lighting patterns.
What was most encouraging was that the signals had a positive effect on people’s trust in self-driving vehicles, with participants reporting that the light signals increased their understanding of the unmanned vehicle’s intent.
Encouraged by the results Ford has put forward a proposal to accelerate the industry coming together to work toward standardization, in parallel with the work the company is already carrying out with the International Organization of Standardization (ISO) and the Society of Automotive Engineers (SAE) to create a unified communication interface for self-driving vehicles.
By so doing the company would like to reach an agreement in three core areas:
- Placement of the signals on a vehicle
- Design of the signals
- The color of the light signals themselves
Ford isn’t the only company working on this communication problem around self-driving cars. Google has its own patent that contains a range of ideas, including light-up “walk” or “don’t walk” signs on the car’s body, image displays, and audible signals similar to the ones used at busy crosswalks.
Drive.ai, a San Francisco-based startup, was working on LED signs on the vehicle that used text and emoji-like pictures to communicate until it was absorbed by Apple’s self-driving car unit.
But what if the vehicle’s paintwork could light up to convey a message?
Electroluminescent paint game-changer
Although electroluminescent paint has found limited support from automotive manufacturers, the novel idea is beginning to draw designers’ interest.
Audi, in the Audi E-tron Sportback concept, has been experimenting with the use of electroluminescent paint on the seats and select door panels to provide ambient cabin lighting. These surfaces gently glow and provide interior lighting without the need for wires, bulbs or even LEDs.
Although a charge is still required for the electroluminescent paint to work, with a coat thickness of 0.02mm, there’s far less packaging required to power it, making it an almost limit-free light source in terms of where to place it.
Darkside Scientific, creator of the world’s first and only patented electroluminescent paint, LumiLor Light Emitting Coating, has been working with European aircraft manufacturer, Airbus, to create electroluminescent exterior markings for aircraft.
The product allows manufacturers to design with light in ways previously impossible. It produces a single-color light that is visible at long distances and shines through many types of atmospheric conditions such as fog, snow and smoke.
According to Vincent Loubiere, lead technologist at Airbus: “This is an exciting new product that can turn surfaces into functional elements in an unconventional way.”
Unlike many light sources such as LEDs, the brightness of LumiLor is distributed across the coating and appears the same from all angles of view. Simply put, anything painted with LumiLor can be turned on and off as easy as a light bulb.
Which despite the many challenges to be met, might make the product ideal for machine-to-pedestrian communication as the number of driverless vehicles on the road increases.
Even as manufacturers work on lighting solutions for fully autonomous driverless cars there is still a requirement for exterior lights to perform their age old function of illuminating the road.
Lighting up the road for driverless cars
“We’re working hard to improve safety in night driving situations,” remarks Mathias Thamm, head of Volkswagen’s Technologies and Innovations department. He says that it’s important because about 30 percent of all accidents involving personal injury occur at night, and accident severity is much higher than during the day.
In automated vehicles the aggregation of data already allows the car to recognize the driving location on a city or a country road, on the highway, or off-road, and will anticipate where other road users might currently be located. The system supports the driver with the best possible lighting without bothering other road users.
For instance, in the Touareg, the LED Matrix Headlamp system performs the LED selection and illumination within the headlamp matrix automatically using Dynamic Light Assist. It turns the LEDs on and off independently and adjusts the lights to the prevailing environment, topography, and traffic conditions.
Core to the headlamp control is a front digital camera that analyzes road conditions, spots oncoming vehicles, and detects traffic signs. This data combined with GPS input, speed, and steering angles determines which LEDs in the matrix will provide the ideal headlamp illumination for the road and surrounding area; in less than a second.
Adaptive matrix headlamps employ a targeted control of up to 128 LEDs to provide precise and optimal light distribution and luminous intensity.
While this technology is designed for current ADAS it is not difficult to imagine its transition to L3 and eventually driverless cars – that is if headlights will be required at all in a world where radar, lidar, infrared cameras and 3D mapping see the world without light.
Irrespective of the detail, there is no doubt that within the next decade automotive lighting will have undergone a quantum shift in design and functionality, adding to the every-growing list of technologies being disrupted by the self-driving car revolution.