Every day, Dr. Narelle Haworth makes the 35-minute commute from her home to her office at the Queensland University of Technology by bicycle. Mostly, she follows established bike paths. She times her commute so that it’s late enough to see and be seen, but not so late that she hits the morning rush hour. And if she can help it, she does not ride on the road.
Every cyclist knows the hazards of riding in traffic, but Haworth knows them more intimately than most. As a scientist at Australia’s Centre for Accident Research and Road Safety in Queensland (CARRS-Q), she has devoted her career to vulnerable road users—pedestrians, bicyclists, and motorcyclists. Her research has convinced local, state, and national governments to take action to protect cyclists. That ranges from speed limits to helmet laws to mandatory three-foot passing rules—all of which have made cycling safer.
But it is still dangerous to be a cyclist. Despite an overall decrease in traffic fatalities, cyclist fatalities caused by drivers have increased in the U.S. over the last 10 years. Haworth’s years of research show that speed limits and passing requirements are effective at preventing cyclist injury and fatalities, but only when enforced. Telling a driver to give a cyclist three feet of clearance is one thing; getting them to actually do it is another. There’s also a robust body of evidence, much of it gathered by Haworth and colleagues, demonstrating exactly what it would take to keep cyclists safe: better traffic design, protected bike lanes, enforcement of speed limits and distracted driving laws, and improved driver education. Yet when Haworth presents these findings, she said she often faces vocal opposition.
Get these f*cking bikes out of the f*cking road.
“It’s hard for cyclists to be taken seriously in any context,” Haworth said. “There’s bad behavior and harassment toward cyclists just about everywhere. In one of our studies of thousands of bicycle riders, what we found was that cyclists would say deliberately being passed too closely was the most common form of harassment. Then there’s a range of other things, including having things thrown at them or drivers cutting them off.”
Animosity toward cyclists often goes off the road and on to social media. Read the comment threads on just about any news story on bike lanes or cyclist safety, and it likely skews more toward anger than support:
“If they don’t want to get hit, bikes need to get out of my way and onto the sidewalk.”
“Cyclists are not policed and think they can do whatever they want. I hate them so much.”
“When I pass a bike with my truck, I like to buzz them. It’s hilarious to see them freak out.”
“Get these f*cking bikes out of the f*cking road.”
Haworth’s research confirmed the obvious: When drivers don’t like cyclists, they don’t respect them. In turn, they don’t respect the laws designed to keep everyone safe. But what wasn’t so obvious is why. Speed limits and traffic laws exist to protect other vulnerable road users too, yet you rarely see drivers bragging about clipping a fourth-grader in a school zone just to show them who’s boss. A funeral procession clogs up traffic, yet no one throws a beer bottle at the hearse. What is it about cyclists, specifically, that fuels such ire from drivers? Haworth suspected unlocking the answer to this question could be helpful in changing attitudes and behaviors.
Cockroaches or Cyclists?
Haworth was thinking out loud about this when a colleague proposed looking at the question through the lens of psychology, not traffic safety. “They knew of someone who was working in the area of dehumanization, or people treating another group as not fully human. There’s research about dehumanization of people of other races, people of other religions, women, all sorts of minority groups, and we wondered, ‘Could this, in some way, be contributing to what’s happening on our roads?’”
To measure this, Haworth and colleagues created an experiment in which 450 people were shown the famous “March of Progress” illustration—depicting evolution from primate to human—or a modified version of the chart that used a cockroach rather than an ape. On the chart, respondents were asked to select where they thought cyclists fit on the ranking. They were also asked to rate how strongly they agreed or disagreed with statements like “I feel like cyclists are open-minded, like they can think clearly about things,” or “I feel like cyclists lack self-restraint, like an animal.”
On both ape-human and cockroach-human scales, 55% of people in the group who hadn’t ridden a bicycle in the last year rated cyclists as not being completely human. Feelings of cyclists as cockroaches or apes corresponded with aggressive behaviors on the road, from shouting and rude gestures to deliberately trying to run cyclists off the road.
In a time when tribalism seems to exist everywhere, from politics to sports rivalries, perhaps it’s not so surprising that “us vs. them” extends to bikes and cars. Dehumanization creates a mental loophole to justify bad behavior: If someone is not human, then surely they cannot feel pain. This disassociation allows people to inflict harm on others (intentionally or unintentionally) with shocking ease. Dehumanization has also been found to increase instrumental violence, where violence is viewed as a means to an end. Drivers with dehumanizing attitudes may find it easier to justify violence against cyclists if it then forces those bikers to a different route, scares them from riding altogether, or otherwise makes it clear who has the power.
“When you’ve got drivers who are expecting to be traveling very fast in a rural area where the speed limits are high and the roads don’t have shoulders, a cyclist is a nuisance,” Haworth said. “Or in an urban area where the traffic is congested and the cyclist is going faster than you are in the bike lane, the cyclist is disrupting the way you think things should be. It’s upsetting, especially if you see them as less than human, less than you.”
A Person Who Rides a Bike
And that sentiment is at the crux of the next generation of road safety. We can build bike lanes and tell cyclists to wear helmets, but the most dramatic impact on road safety will come from getting drivers to see their actions as more than just crushing a cockroach under their wheel.
“So now we need to figure out how to make cyclists seem more human again. Some say we should stop using the word ‘cyclist’ and instead say ‘a person who rides a bike’ so that we put the emphasis that it is a person doing the activity, not another species,” Haworth said. “It does seem to be associated with quite a different response in people’s minds. People are more positive towards bicycle riding than they are toward cyclists, particularly cyclists wearing lycra.”
Experts have also pushed for changes in the way people talk about vehicle-cyclist interactions, by replacing “accident” with “crash” and switching from passive to active voice. “Cyclist hit by car” is detached and underplays the driver’s actions or choices. Instead, “Driver hits, kills cyclist” is a stark, yet effective reminder of the humanity present on the road. Public-messaging initiatives like It Could Be Me, which aims to re-humanize cyclists by putting faces and stories to the people sharing the road, may also be a tool in increasing driver empathy. When you see someone as a fellow human being, with a beating heart, a soul, and a family, you’re less likely to want to harm them.
But changing minds is easier said than done, especially right now. The notion of sharing the road is difficult when humans can’t even share a common set of facts. Which begs the question: What if we removed humans from the equation altogether?
For more than 50 years, the Consumer Electronics show, or CES, has been the global stage for innovation. The mega-sized industry trade show is a real-life version of the 1960s cartoon The Jetsons, where robots, flying cars, and smart homes graduate from science-fiction novelty to real-world presence.
At the 2018 show in Las Vegas, in a conference room tucked between the debut of a 3D camera and a nightclub featuring robot strippers, engineers from Ford Motor Company, Trek Bicycle Corporation, and Tome Software were breathing life into a science-fiction novelty of their own: a way for bikes and vehicles to talk to each other, absent of the humans steering them.
The three companies came together to discuss C-V2X, a framework which enables vehicles to communicate with each other and the surrounding environment, like stoplights and construction zones. C-V2X (which stands for Connected Vehicles to Everything) uses Bluetooth and 5G wireless technology to broadcast a device’s location, speed, acceleration and other elements while simultaneously listening to other devices; if one car makes a sudden stop, the other connected cars behind will follow suit, avoiding a rear-end collision. Would it be possible to leverage this technology to protect cyclists as well?
Even all companies being in the same room was a breakthrough in and of itself. Despite sharing the same road, the American bike and auto industries don’t have much of a collaborative history. Safety is a priority for both, but in different ways. Automakers are eager to get a high safety rating from the National Highway Traffic Safety Administration (NHTSA), which only assesses how well the car protects the people contained inside—not pedestrians, cyclists, or those in other cars. The incentive, then, is for automakers to design bigger, taller, heavier SUVs and trucks, which now dominate the market. In 2019, more than two out of three new vehicles sold in the U.S. were SUVs, pickup trucks, and minivans. Of those, 150,000 were too big to fit into a standard parking space. By 2025, these large vehicles are expected to account for 78% of new vehicle sales.
For the cycling industry, the focus is on shielding people on bikes from supersized rides. Vehicles may have become safer for people in cars, but people on bikes have paid the price. Because of their size, SUVs and trucks are more deadly than compact or midsize cars. Heavier weights mean more force upon impact, and the taller height of the vehicles makes it more likely that impact on a cyclist will take place on the head or torso instead of the legs.
Because of these different perspectives, the vehicle and bike industries have historically held a similar regard for each other as drivers and cyclists. But that all changed with C-V2X. Suddenly, the conversation became less about hostility and blame and more about actually sharing the road.
“With C-V2X, now we can talk about how do we manage traffic flow more efficiently? How do you manage pedestrian safety more efficiently?” Eric Bjorling of Trek said. “That’s when we started to think maybe we could make something happen together. Within weeks, we had most of the large bike and component manufacturers, automotive manufacturers, all on board with developing this technology.”
A New Language
The bridge between the two industries was Jake Sigal, founder and CEO of Tome Software. In 2013 Ford bought his first start-up, Livio, which connected mobile phone apps to cars. A year later, Sigal founded Tome. He continued to work on automotive software, but his passion for bikes led him to work on smartphone apps and technologies for cyclists as well. It was an organic next step, then, to build a platform for bikes and vehicles to connect and communicate with each other.
“In a way, it’s a language,” Sigal said. “Like the analog approach today, if you have a bike bell and ring it, cyclists, drivers, bus drivers, pedestrians, they hear the bike bell and it’s universal. They know what that is and that they need to respond. With this, the car ‘hears’ the bell and knows how to respond.”
Navigation systems in vehicles, cycling equipment, and smartphone apps already provide the ability to send or receive alerts and messages, so allowing bikes and vehicles to communicate is quite feasible. Using Artificial Intelligence (AI), bike-to-vehicle technology can alert the driver to a moving bike in the path of the vehicle or potentially dangerous conditions, such as a cyclist intending to turn or suddenly slowing down. This early alert makes the driver look up and take corrective action if needed. The goal of these alerts is to nudge drivers into making safer decisions—just as a seat belt chime prompts everyone in the car to buckle up, the cyclist alert could persist until corrective action is taken. To avoid the problem of over-alerting—for example, a chime when passing a coffee shop with bikes parked outside—programming would account for whether the bike is in motion, how far it is from the vehicle, and how fast it’s traveling.
Different ideas and different interpretations, but all centered around the same goal.
For this to work, a two-sided business model is essential. It needs to be easy for millions of bikes and millions of cars to become equipped with the technology. To expedite this process in the name of cyclist safety, Trek, Ford and Tome are moving the technology behind bike-to-vehicle communication to a consortium focused on industry standards. In addition to speeding up the development process, creating standards will increase the number of bikes and cars sharing information with each other, making roads safer for all.
“If this is a thing where only Trek bikes can communicate with only Ford vehicles, it’s not going to work,” said Tony White, principal engineer at Trek. “If we’re going to create something that is widely adopted, we have to make sure everyone is speaking the same language.”
Industry standards also mean new ideas can fit into already-existing technology. “The beautiful thing about this project is that you have different companies looking at different solutions,” Bjorling said. “We could put this into new bikes, but that doesn’t address the millions and millions of bikes that are already on the road. Now we’re building something that can retroactively be put on any bike. Trek is building it into an aftermarket light, but somebody else might say, ‘Well we’ve got a ride tracking app on our phone, we could just put this into the next update for the app.’ Different ideas and different interpretations, but all centered around the same goal.”
“It’s so interesting to see how others think about the same idea,” White said. “You think you’ve got this one great idea, and then when other people see it, they start to say, ‘How can we use this in another way?’ It snowballs and grows into something bigger. [When I first started designing and testing bike lights] I had no idea what it would become. But somebody said, ‘What if you just add a little bit of electronics on your bike?’ And one innovation for cyclist safety turned into another.”
What About the Trolley Problem?
Partnerships between the bike and car industries now also set the stage for a continued collaboration as the auto industry evolves toward autonomous cars. In a way, the notion of driverless vehicles seems like a boon for people on bikes. After all, cars don’t experience emotions, much less hatred toward cyclists. They can’t be distracted by a text message. They follow the rules of the road, because rules are all they understand. Technology is impersonal and detached.
And that might create a whole new set of problems for cyclist safety.
In the classic thought exercise known as “The Trolley Problem,” a fictional scenario is presented in which a trolley driver careens down a track, only to discover the brakes of the train are no longer functioning. The driver is presented with two options: continue down the current track, where five people are tied up, or pull a lever that will divert the trolley to another track, where one person stands. Do nothing, and the trolley kills five. Pull the lever, and one person dies. Which is the most ethical choice?
Now consider a modern-day version, in which a cyclist suddenly swerves out of a bike lane and into vehicular traffic to avoid colliding with a driver opening the door of a parked car. Should the autonomous car be programmed to swerve into oncoming traffic to spare the cyclist, or protect the passengers by going straight?
“It’s something that comes up on a regular basis, the trolley problem,” Sigal said. The MIT Moral Machine poses a number of hypothetical scenarios for autonomous vehicles, each one more ethically cloudy than the last. These scenarios remind us that even though a fully autonomous vehicle will one day be capable of performing all driving functions under all conditions, humans and human judgement will still be baked in. Autonomous vehicles cannot think or make decisions, but instead they carry out the tasks they are programmed to do. That programming is done by humans, and humans have biases and blind spots (like developing facial recognition software that struggles to recognize non-white people).
The vehicles also have to respond to unpredictable events from humans, animals, and the environment around them. Take, for example, the 2018 death of Elaine Herzberg, who was crossing the road in Tempe, Arizona, when a self-driving Uber vehicle hit and killed her. Even though the vehicle was equipped with collision avoidance measures, it wasn’t programmed to expect pedestrians crossing the street outside of crosswalks. Further, the test driver behind the wheel, Rafaela Vasquez, was supposed to override the self-driving vehicle in an emergency. Instead, she was watching videos on her smartphone.
Enemy of the “Good Enough”
Autonomous vehicles will likely reduce the number of injuries and fatalities caused by cars, but it isn’t a perfect solution. There will be growing pains as the technology evolves from no automation to full automation. In the interim, cars of varying levels of automation will be sharing the road. It will also be years before the majority of people on bikes are equipped with C-V2X technology. The transition to a fully connected body of transportation will not be a seamless one. That’s why transportation researchers and engineers insist technology is not a substitute for human education and diligence.
“We prioritize technology that addresses vulnerable situations when everybody’s following the law,” Sigal said. “So if a cyclist is blowing an intersection, or cars are blowing an intersection, that’s a problem. We prefer to first start off with drivers that are sober and drivers that are not distracted. So our priority is when everybody’s following the law, and when there’s a situation where we can help reduce the likelihood of a collision.”
But people don’t always follow a predictable, law-abiding pattern, so when faced with problems of the trolley/cyclist variety, engineers are tasked with anticipating all possible outcomes and finding the best possible solution. Instead of choosing between prioritizing the driver’s life or the cyclist’s life, engineers are tasked with minimizing damage overall.
“So let’s say, for example, that a cyclist is riding a bike, they’re in a bike lane, and then all of a sudden a door opens from a parked car into the bike lane. The cyclist immediately moves into the next lane over in front of a vehicle. The vehicle then slams on the brakes,” Sigal said. “Because what is killing cyclists is the speed of impact, not the impact itself. Impacts are bad, but the data shows that when a cyclist is hit at over 35 mph speed, the death rate just shoots up. So even if there is impact, if we can reduce the speed of the impact by emergency braking, that allows us to increase the chances that you don’t lose a loved one that was out there.”
It’s not a perfect solution, Sigal admits, but it’s the right place to start. If engineers were to get caught up in perfection, they’d never move forward on improving bicycle safety. “We don’t have to solve the trolley problem to release technology. There are so many what-ifs that could stop us, but we’ve got to keep doing what we’re doing. We’ve also got to try to create those situations where we can provide critical informing messages to humans or computers.”
That means returning to the core of cyclist safety: acknowledging that people on bikes are, in fact, people. And paradoxically, technology may be able to do just that.
“As a cyclist myself, technology needs to help humanize cyclists to reduce road rage,” Sigal said. “I think changing the way cyclists are humanized comes down to communication. If I’m on a bike, and I pull up to a stop sign, I’m gesturing to say ‘Go ahead,’ or I give a thumbs up to say thanks for giving me a little room, I appreciate that. There’s a lot of these cycling signals that we use with other humans, and I would love to be able to digitally say thanks to all the good drivers with technology as well.”