- Novel laser-based system enables autonomous cars to predict what’s happening ahead
- Can self-driving cars put an end to drive-by shootings?
- Autonomous vehicles can even detect signs of COVID-19
- Self-driving cars will communicate with traffic lights and other road users
- New perception method vastly improves obstacle detection in autonomous cars
- Will the superpowers of self-driving cars prevent crimes and save lives?
Although we’re still far away from fully autonomous cars with AI-powered brains under their hoods, these nifty pieces of technology can already do amazing things. For one, today’s self-driving vehicles are much better and safer drivers than we are – a rather significant fact, seeing that human error is the cause of a staggering 95 per cent of all accidents in the EU. In 2017 alone, traffic accidents claimed over 25,000 human lives, which is one of the main challenges facing the transport sector. Fortunately, thanks to the latest developments in cutting-edge technologies like intelligent algorithms, we’re inching closer to autonomous vehicles whose performance exceeds our wildest imagination. And the automated transport sector is thriving. In the EU, for instance, the self-driving vehicle market is expected to reach revenues of up to €620 billion by 2025. Aside from contributing to greater road safety, in the future, these clever vehicles will be able to do a whole host of other incredible things as well.
Novel laser-based system enables autonomous cars to predict what’s happening ahead
As we’ve written before, LiDAR (Light Detection And Ranging) allows autonomous driving systems to ‘see’ the road and avoid potentially dangerous situations and fatal accidents. LiDAR has come a long way; from being rather slow at first, to easily handling motorway speeds. A Stanford research team has recently come up with a novel laser-based system that has the ability to even reproduce images of hidden objects, allowing self-driving vehicles to see objects that are around a corner. “There is this preconceived notion that you can’t image objects that aren’t already directly visible to the camera – and we have found ways to get around these types of limiting situations,” said Dr Matthew O’Toole, a coauthor of the research. Dr. Anna Anund, who is leading an EU-funded project focused on developing software for driverless car sensor systems, underlined the importance of this breakthrough. “The automated vehicles need to be able to predict or detect what is happening ahead. This is not only what is going on in [their] own lane but also in the surroundings like pedestrian paths, on the other side of the corner etc.”
Can self-driving cars put an end to drive-by shootings?
Apart from car accidents, which will still be leading to casualties for a long time to come, some communities are also dealing with other vehicle-related incidents, such as drive-by shootings – often deadly acts of violence committed by gangs and terrorists in various parts of the world. Not too long ago, for instance, CCN reported that, “fifteen people were shot in a drive-by shooting in Chicago at a funeral for a victim of another drive-by shooting,” according to the Chicago Police. “We can’t keep meting out violence with violence,” emphasised superintendent David Brown. “An eye for an eye makes us both blind. It’s destroying our families and perpetuates this endless cycle of gunshot victims night after night.” Similar incidents happen in the EU as well. In 2019, for instance, Zelimkhan Khangoshvili, an ethnic Chechen holding a Georgian passport, was killed in a drive-by shooting in a suspected act of Russian state terrorism. The biggest problem is that law enforcement is unable to prevent these crimes and such acts need a more comprehensive approach.
In today’s ‘dumb’ vehicles, perpetrators of drive-bys can easily be hidden from sight. Self-driving vehicles, assisted by better infrastructure and greater connectedness and equipped with inward-facing cameras and an audio system will be able to ‘see’ and ‘hear’ what goes on inside, as well as outside. Despite the obvious privacy concerns this may cause, this tech could, on the other hand, help law enforcement uncover those with malicious intentions and prevent drive-bys and other crimes. The data collected by these technologies can be used to detect firearm shootings from within the vehicle and pinpoint the exact location of the car. In fact, not a single thing within these super-intelligent cars will go unnoticed, which will make most people think twice before committing a crime.
Autonomous vehicles can even detect signs of COVID-19
Since the outbreak of the novel coronavirus, many companies have been fiercely dedicated to finding the best way to catch the disease in its earliest stage. And soon, our smart cars – yes, you read that right – might know how to recognise COVID-19 symptoms better than we or our thermometer can. Since 2015, the Israeli company Foresight Autonomous Holdings Ltd. has been working on the development and implementation of sensor systems for the automotive industry. But now, they have another mission – the design of a multiple automotive vision system to help fight COVID-19.
“The current thermal-screening solutions only detect elevated body temperature. Taking into consideration that lack of initial fever is common in COVID-19 cases, patients with no fever may be missed if the surveillance definition focuses on fever detection only,” explains company founder and CEO Haim Siboni. “The use of both thermal and visible light cameras allows us to detect several symptoms, in addition to elevated body temperature. This wouldn’t have been possible by using thermal cameras alone.” Implemented in self-driving vehicles, this intelligent system would be capable of combining detected symptoms, increasing the likelihood of accurate illness detection before more severe symptoms appear. Timely detection and action can save lives, and that’s where cutting-edge tech can help us prevent developing corona-related life-threatening conditions.
Self-driving vehicles will communicate with traffic lights and other road users
Human drivers and pedestrians often use non-verbal signals, including facial expressions, hand gestures, and eye contact, to communicate in traffic. While autonomous vehicles can’t communicate in the same way, these vehicles do have the ability to engage with pedestrians through explicit and implicit means. According to a study conducted by researchers at the University of Michigan, this explicit communication consists of using light-emitting diode message boards and interactive headlamps. While this opens up many exciting possibilities, these developments also have their downsides. For instance, if there are too many autonomous vehicles on the road, this can lead to information overload and confuse pedestrians and other road users. Implicit vehicle communication is a less explored approach and refers to autonomous vehicle behaviour cues, such as its motion and kinematics.
Researchers also explored the use of traffic lights in human-autonomous vehicle interactions. Their study, which involved 30 participants, reveals that pedestrian trust in autonomous vehicles is influenced by the presence of a traffic signal. Since traffic lights can be considered a higher form of authority, self-driving vehicles are expected to obey these signals. Therefore, participants in the study showed higher trust in self-driving vehicles at intersections with traffic lights. Further advances in the field would make it possible for autonomous vehicles to communicate with traffic lights even before these become visible to the vehicle. For example, on a slippery road where an abrupt stop could be dangerous, the vehicle could communicate with the traffic signal through a wireless network and ask for a longer green light.
New perception method vastly improves obstacle detection in autonomous cars
Advanced pedestrian and vehicle recognition systems are what makes driverless cars safer and more efficient than traditional vehicles. However, to better detect objects and people on the road, autonomous cars need to be able to detect empty space as well. Researchers from Carnegie Mellon University (CMU) have shown that this is possible and presented their findings at the Computer Vision and Pattern Recognition (CVPR) conference, held in June 2020. As part of their research, the team used map-making techniques to help the self-driving car’s perception system “reason about visibility when trying to recognise objects.” This is similar to a digital model some companies use to detect when certain office spaces are occupied or empty. “But that doesn’t always occur for live, on-the-fly processing of obstacles moving at traffic speeds,” says Deva Ramanan, an associate professor of robotics at CMU. The method, which takes only 24 milliseconds to run, improved the vehicle’s detection abilities by 5.3 per cent for pedestrians, 7.4 per cent for trucks, 10.7 per cent for cars, and 18.4 per cent for buses. In comparison, LiDAR technology takes 100 milliseconds to map its surroundings using 3D information.
Will the superpowers of self-driving cars prevent crimes and save lives?
Scientists and engineers are working hard to develop technologies that will make self-driving cars safer. But more often than not, new technological breakthroughs can even exceed our expectations. Decades of research have resulted in autonomous vehicles that can see, hear, and even sense their surroundings. Today, smart cars can communicate with traffic lights, detect weapons and malicious intentions, and even pick up the first signs of COVID-19. And despite concerns surrounding self-driving systems, the benefits still seem to far outweigh the potential downsides. We suggest you buckle up and enjoy the ride, as the cars of the future seem to be turning out to be heroes with real superpowers.