“Safely” Exposing People to Danger:
Using Accident Videos to Raise Risk Awareness

Kan Shimazaki
Assistant Professor, Faculty of Human Sciences, Waseda University

John McClane is always out of luck, so once you’ve seen the third or fourth Die Hard movie, you can pretty much tell when he’s going to run into some trouble. Before Die Hard, though, all we saw were action heroes rushing in to save the day—rarely were there scenes where an unarmed, off-duty policeman got caught up in something. So when the first Die Hard came out, its unexpected storyline must have kept you on the edge of our seats the whole way through.

When we repeatedly see or hear something with a recognizable pattern, we start being able to make predictions about things that exhibit similar patterns. When it comes to entertainment, we can get enjoyment from the fact that things are unpredictable; in the real world, however, unexpected events often bring problems and inconveniences. What this means is that we need to use our past experience with certain patterns to predict when something bad is likely to happen. This way, we can prepare ourselves to handle what’s coming—and in doing so either prevent an accident entirely or at least minimize the damage.

Figure 1

The same thing happens when we drive a vehicle. Through repeated exposure to danger, we eventually learn make predictions, thinking that “Someone always pops out from that blind corner,” “This is a classic setup for a motorcycle accident,” and so on. As we do so, we gradually become safer drivers, and less likely to get involved in an accident (see Figure 1). That said, for better or for worse, the frequency with which we encounter risky situations while driving is fairly low. Japan requires its drivers to log about 20 hours of on-road training time in order to get a license, but rarely do people actually experience danger during those hours. And luckily (unluckily?), even if they do, there are so many different accident patterns that there’s no way for us to experience them all. Finally, because risky situations can easily lead to accidents, it’s probably best that we avoid exposing people to danger in the real world.


Figure 2

The driving simulator (DS) was developed as a means of “safely” exposing people to danger. But not only is a DS far too large to put in someone’s house, the cheapest one available is more expensive than a car itself. Also, because the DS machines used for educational purposes only offer a handful of preloaded driving scenarios, they tend to get boring after a few uses. Not to mention that Japan has some 80 million drivers, and there are nowhere near enough driving simulators to offer repeated training to them all.

When it comes to a factory, medical, or aircraft accident, there are almost always at least a few minutes between the time a risky situation emerges and the point where it becomes critical. Traffic accidents, on the other hand, leave at most a few seconds between the time you notice the other person and the moment of impact. This is not long enough for a person to respond relying solely on a rational understanding of the accident pattern. It requires developing a kind of “knack” for identifying them—and because developing this knack is a matter of reflexively turning one’s eyes towards dangerous areas without almost no conscious thought, as shown in Figure 2, it must be developed through repeated practice. It’s a lot like learning to play a sport or a musical instrument.

Figure 3

Driving simulators have their problems, but if all we need to do is show different traffic situations and get people to learn which patterns are likely to lead to an accident, we don’t need fancy equipment like a steering wheel and pedals, a 360º screen, and a six-axis motion control platform. Instead, our research group has developed a tool that uses the iPad to teach people to recognize the patterns most likely to lead to an automobile accident (Figure 3). The training images are not computer-generated graphics like the ones used in driving simulators; instead, they are actual accident images taken from in-vehicle drive recorders. Trainees simply tap any areas of danger they see as they watch the video on the screen. The software records whether users tap the potential collision targets (or blind spots where collision targets may be hiding) and then provides feedback on whether they noticed the dangers, based on the timing of the tap and the like. At the end, users watch the actual collision footage—which makes it absolutely clear where the risk of collision was. Because accident footage is collected on drive recorders on an almost daily basis, new videos can be distributed online as they become available. This ensures that users never get bored with the program.

We have subjected our training tool to multiple effectiveness tests, and have confirmed that it does yield beneficial effects in terms of getting people to look in the right places when driving, motivate them to be more careful drivers, and so on. It’s even starting to be adopted by taxi and insurance companies. We are currently working to add greater functionality to the program, such as linking it to hazard maps and adding leaderboard displays, so that people even more engaged in the training. The software can be downloaded for free from the App Store*, so please get your copy and spread the word. Every time someone does the training, we get one step closer to reducing traffic accidents.

The app is called HazardTouch. Please note that accident videos are only available to a limited number of users at this time.

Kan Shimazaki
Assistant Professor, Faculty of Human Sciences, Waseda University

Kan Shimazaki graduated from the School of International Relations, University of Shizuoka in 1999. After working as a professional driver of taxis and large trailers, Shimazaki decided he wanted to study traffic psychology, and asked to study under the Waseda University professor and former chair of the Japanese Association of Traffic Psychology Toshiro Ishida. Shimazaki completed his master’s and doctoral programs in the Graduate School of Human Sciences, Waseda University, and has taught as an assistant professor at the university’s Faculty of Human Sciences since 2011. In addition to holding all of the first-class driver’s licenses, Shimazaki also holds a number of other certifications that allow him to operate heavy equipment, handle hazardous materials, and more. He is qualified as a Chief Traffic Psychologist and a JSAE Engineer in addition to holding his doctorate in human sciences.