The Future of AVs Depends on Non-Stop Vision, Day & Night

While the world sees an acceleration in AV testing, technology must still overcome the wrath of Mother Nature

‍

For years, the world has been edging closer to a future that sees autonomous vehicles (AVs) on our streets. Whether it’s taking employees to an office or picking up kids from school, there’s no doubt that the expectations set by tech giants like Tesla are keeping excitement strong for when that day will be.

The fact is, there are problems with AVs in their current state that companies won’t yet admit or that eager-eyed tech fans don’t see. Having a vehicle operate with full autonomy day and night in any weather is a risk that vehicle manufacturers cannot gamble - and rightfully so if lives depend on it. While autonomous vehicles are safer than human drivers, they still can’t operate in all lighting and weather conditions. They must overcome unpredictability when faced with night driving, adverse weather, limited visibility, and small objects on the road.

These objects can be identified with GatedVision technology that works in a completely different way from any technology out there. By placing a small camera behind the rear-view mirror and a near-infrared illuminator incorporated into the headlamps, vehicles can identify objects at further distances and nighttime fogginess. While daytime driving is pretty solved as a path forward, nighttime and all-weather driving is still holding the industry back. At least, until now. 

In this article, we’ll dive into the future of autonomous vehicles, the significance and barriers of 24/7 functionality, and how Bright Way Vision’s GatedVision technology helps overcome these barriers.

What Are Autonomous Vehicles (AVs)?

To put it simply, autonomous vehicles are cars, trucks, trains, or industrial vehicles that can operate with little (ADAS) or no human intervention. There have been varying degrees of autonomy in the transportation sector for decades, but recent attention has been given to reaching Levels 4 and 5 - where a car can start, navigate, park, and drive without human intervention. 

There are many technologies that car developers like Tesla or Waymo use to make sure their vehicles drive on roads without help, working in tandem with a standard camera (mimicking the human eye spectrum) that's good for daytime and good visibility conditions, but they’re limited to low beam ranges at night. Some common ones are LiDAR and radar - technologies that use cameras and sensors to scan the road and spot traffic lights, pedestrians, or threats using deep-learning algorithms. Furthermore, we should take into account that two-thirds (65%) of the year is categorized as ‘visually challenging’ due to it being night time, and combined with low visibility weather conditions (fog, rain etc.,) and glare, which increases the number significantly, therefore these technologies are restricted to regions with consistent and predictable climates such as the deserts of Arizona or Israel. 

The Future of AV and the Current Pace of Adoption

Elon Musk once predicted in 2015 that Tesla is very close to Level 5 autonomy. Two years later, he made the prediction again - and yet there are still hurdles that every AV manufacturer needs to overcome to reach this self-sufficiency. The majority of experts claim that technologies like LiDAR are essential for the success of AVs, and yet Musk again called users of the technology “doomed”.

The prediction for 2020 itself was that we would have around 10 million AV’s being tested, as of now in the US there are over 1,400 registered AV’s. And the next future prediction for 2030, quoted by Statistica, we are expecting to see 1 in every 10 vehicles will be autonomous. So it seems predictions did not meet expectations, how likely can the next forecast be met?

The Significance of 24/7 AV Operations and Barriers

• The public perception of autonomy
  Even though 94% of road accidents are caused by human error, a staggering 71% of Americans have said they won’t ride in a fully autonomous vehicle. The challenge for companies today isn’t only working to improve the technologies of laser, LiDAR, and computer vision, but companies need to get the public on board with their view of future mobility. 

• Achieving trust in AV’s through uncertainty
  Achieving trust in AVs will seldom be achieved if over 65% of driving conditions is unsuitable for them to operate. Small towns in countries and states with clear weather are seeing fleets on their roads, but the unpredictability of weather is forcing these technologies to face an uncomfortable truth - they’re not yet enough, especially considering that driving at night is unsolved.

• Driving in all seasons
  It’s not good enough that a car with current autonomous capabilities can identify threats only on a road in certain conditions. For trust to be built and autonomy to be achieved, the vehicles need to face unknown challenges and a myriad of environments. The experience for the passenger needs to be the same in the morning and night; in winter or in summer. This can be achieved by VISDOM, a camera system with GatedVision technology from Bright Way Vision.
  A WIRED article provides a breakdown regarding some of the most challenging conditions being - fog, snow, or black ice on the roads left undetected. Current sensors do not have the capabilities to tackle this, nor is AV testing conducted during these non-ideal conditions. What they need is a way to combine these sensors into one suite that can extend and clarify the images currently being recorded, not to mention the ability to have strong AEB (automatic emergency braking) capabilities, which are a key element of accidents and fatalities.

How Bright Way Vision’s GatedVision Technology Overcomes Barriers

Level 5 autonomy can never fully be implemented in fleets or taxis until vehicles are prepared to drive in any and all weather, especially at night. 

Bright Way Vision provides non-stop, long-range, and robust vision clarity for autonomous vehicles regardless of the weather in which they’re traveling. Regardless of the lane, speed, or new warning signs, vehicles will be able to work 24/7 in all weather and lighting conditions - something particularly important for truck fleets that will transport goods at any time of day. 

Initial research has shown that GatedVision, which accumulates multiple exposures in a single camera frame for its sensory input, can uniform lighting and ultimately extend the visibility of the gated camera at nighttime, in all weather conditions. This provides the key system that AVs rely on, thus becoming the ‘eyes’ that are so critical for each vehicle.

The all-weather camera system, VISDOM, provides GatedVision technology for taxi fleets, trucks, and robotaxis designed to transport passengers in a variety of weather. When cars can see farther during their journeys, they have more time and context to identify threats and analyze important next steps to ensure the safety and comfort of passengers and pedestrians.

When this is achieved, it will be because AVs have a suite of technologies that make them safer, smarter, and more intuitive to new challenges. Bright Way Vision predicts vehicles will contain cameras with radar and GatedVision combined to cover any situation. The cameras, unlike today, will have no blind spots and cover the full spectrum of what’s in front of them. Notably, the system will cater to each vehicle for its need - whether it’s a truck going non-stop for days or a subway carrying people to work.

Conclusions

The roads might be paved with good intentions, but they must overcome technological and cultural challenges first to successfully deploy gated imaging technology in vehicles. As the public learns to trust the transformative mobility solutions that will reach our roads in upcoming years, it is essential that cars must be ready to meet the challenges ahead. 

LiDAR, far infrared, deep learning, and other technologies are not enough. While these technologies are ripe, they’re not qualified enough for the automotive sector due to their price and partial results. They must be consolidated by taking a view of the road and sending a signal to the electronic control unit (ECU) which controls the car. Only this can provide the technology to see into the roads, and the future, of autonomous vehicles.