Safe roads of the future

Intercity single-lane roads are the most dangerous for drivers and have always posed a threat to people's lives. These are the very roads where every 10–15 kilometers you can see artificial flowers tied to a tree, a road sign, or a small monument. The combined speed of oncoming and following traffic is over 200 km/h, which is deadly in a collision. Among the already known ways to improve road safety, the following can be highlighted (not taking into account the speed limit approach, which doesn't work, just as installing additional cameras is also not an effective solution over the entire distance):

1. The simplest thing is to make two lanes in each direction. An expensive option, often requiring demolition of residential and other structures, reworking bridges, draining marshy areas, and many other complex works. Therefore, in this direction, as a rule, no work is carried out at all on already open roads. Although it would be possible to temporarily introduce a second lane for several kilometers (alternating in different directions) where it can really be done cheaply, and leave single-lane sections near populated areas, bridges, and other difficult stretches. In some countries, this practice is quite common.
2. Where it is impossible to make two lanes, there is a good practice (including in my country) — to make wider shoulders and also pave them with the same asphalt as the main part of the road. In that case, the actual width of the road is enough for 3 or even 4 cars to fit in a row. If, in addition to this, we instill a driving manner where drivers ahead slightly move to the side and let you pass, this will significantly reduce the risk of a possible head-on collision, since you won't need to pull into oncoming traffic. The downsides: you need to be careful with such wide single-lane roads near populated areas, where people may be walking on the shoulder, and also that not all single-lane roads can be covered this way.

Now we can go a bit further and think about what other innovations could be devised to make traffic on such roads safer.

1. A technology that today seems to exist but is extremely rare — is displaying on large vehicles from the rear, in the form of a screen, front-view cameras that would broadcast the road ahead of the vehicle. I think 99.999% of trucks in the world do not have such functionality, and it will never be installed on current vehicles. So, most likely, this approach is doomed to fail from the start, although in certain situations it could be beneficial and save someone's life.
2. Using drones to monitor violations. A technology that, unlike the previous one, is already being used and is aimed more at punishing committed offenses rather than reducing danger on the roads. You can't assign a drone to every road, to every stretch, and besides, as I already said, it won't save you from the accident itself, but can only capture it or issue a fine to the violator if they survive the overtaking.

Installing cameras on heavy trucks, as described in point 1, will not solve the problem on curves, ascents, and descents, where daredevils will still rush and break the rules, and drones will not make overtaking safe for those who decide to attempt it. In such cases, artificial intelligence can help us. And here there may be several options.

1. Smart cars. The essence of the idea is that manufacturers of all new cars would set as a standard some GPS device or radio transmitter that would allow all vehicles within a radius of, say, 2 kilometers to identify themselves on the road and read other participants. Moreover, when attempting to overtake, this device, operating on the position, distance, and speed of other vehicles, would clearly signal whether it is possible to overtake now or not. Most likely, this device would not be a panacea, because: firstly, even if car manufacturers agree and start mass production of cars with this device, it will take about 50–60 years before all current cars cease to be used (on the other hand, cars will still be means of transportation for hundreds of years, and maybe even thousands, and 50 years is not that much). Secondly, this is a potential vector for hackers to track a specific car or group of vehicles. Thirdly, such devices can also fail, and even linking the serviceability of this device as a mandatory item for technical inspection will not always work (unless you explicitly embed into the vehicle's logic a condition: not to operate and not to drive if the device is not functioning). Nevertheless, despite all the drawbacks and implementation difficulties, the advantages of such devices are significantly greater, and in the long run, they should improve road safety.
2. Smart roads. If in the previous point the role in deciding whether to overtake or refuse is performed by the car by receiving information from other cars, then in this case, such a role would be given to the road. The concept involves installing on every section of the road micro-cameras or, rather, micro-sensors that could read all moving vehicles on their own and neighboring sections (by exchanging information with other sensors). Such sensors should be placed very close, perhaps every 20–50 meters. Plus, they should ultimately communicate their decision about overtaking permission to the cars, so vehicles would also need some devices (though simpler than in the previous point). Among the advantages of this approach, it can be noted that the roads will be able to analyze all moving vehicles, including those that do not have special receivers-analyzers for overtaking safety, and notify all nearby traffic participants about such vehicles. A second obvious advantage of this approach could be the speed of the algorithm and decision-making, since now we don't necessarily need GPS connection or radio waves for communication between the devices themselves, installed every 20–50 meters. Such devices could be connected by a physical cable, which might be more reliable than radio waves. Among the disadvantages, as I already noted: besides communication between the devices of the road itself, you also need to support the need for additional communication with vehicles. And with cars, you can't use a physical cable anymore, though I don't think that's a problem. The second downside is that there would be a need for a great many of these road devices, and they must be very reliable, because if they frequently break down, a whole mobile road crew would be needed, dealing only with repairing and replacing sensors. Moreover, these sensors and the analysis system will be introduced gradually, and on smaller roads, they will be scaled significantly later than on busy highways. Thirdly, the security issue hasn't been canceled, and hacking such a system is dangerous due to the spread of false information to vehicles.