The main idea
In the event that a policeman is attacked, this should be reported automatically to the central console. This system can save the life of a policeman, if he is alone and a wound has led to a loss of consciousness. Under normal conditions, a policeman is threatened with death due to a large loss of blood. The new system will not only report the fact of the attack, but also the location of the policeman, the position of the wound on the body, his pulse, body temperature, etc. She also will tell if the bullet wound is through or the bullet is stuck in the body. Moreover, not only the number of bullets will be reported, but also their caliber (here at present only a rough estimate is possible). Theoretically, you can even determine the number of attackers. Photographs and the recording of sound from the crime scene will also be made and sent. The system distinguishes between knife and bullet wounds.
The system will also notify the post about the incident not only if the policeman was shot, but also if the policeman shot himself.
The system will consist of several subsystems. Each of them is independent of other components and provides quite a bit of information. But combining the data from several sources, we can make quite interesting and important conclusions about what is happening or was happening to the policeman.
So, the system consists of the following components:
- Body armor
- Pistol and holster
- The Wrist Tracker
The first system is embedded in a standard police bulletproof vest. Its task is to report whether the policeman was fired, and if so, to what consequences this resulted. At the moment, several approaches are being tested to solve this problem.
As already mentioned, the technical details are unknown. We can only talk about common ideas.
So, the first approach to solving the problem. Between the outer and inner layers of matter and armored plates there is a layer of metal mesh.
In case of damage, the system will accurately determine the place of breakdown of the armored plate "ringing" the grid. If the outer layer of the grid was hurt, then there was contact with the bullet.
If there is also damage on the inner layer, then the bullet passed through the armored plate and, accordingly, one can speak of a gunshot wound. In case there is damage to the inner mesh on the other side of the body, the wound is through. Having determined the difference in time, when gaps formed, you can determine the direction of the flight of the bullet. For example, if there was a rupture on the back, and then on the chest, then they shot from the back. Moreover, knowing the position of the ruptures, you can determine the trajectory of the bullet in the body and thus predict possible damage to the internal organs.
This solution works well for firearms, because It has great energy and is guaranteed to damage the grid. In the case of stabbing, the system may not get enough damage to reliably determine the attack. Yes, the knife will not break the bulletproof vest, but the attack must be reported, so the system should reliably recognize such cases.
The second approach greatly improves the ability of the system to recognize an attack with low-energy weapons, for example with a knife. Under the upper layer is a layer consisting of two components
- a layer of polyethylene with fine bubbles of liquid (in the figure a layer of yellow color)
- a layer of material that, when wet, begins to conduct electricity (in the figure a layer of red color).
The principle of operation is very simple. The knife is pierced by a bubble of water, water moistens the adjacent layer, which begins to conduct electricity.
The principle of the layer that determines the place of "leakage" from the liquid layer, unfortunately, is not explained.
Another option takes into account the fact that As a rule, bullets and knives are made of metal. At the time of contact, the total capacity of the system should jump-jump. By measuring the capacitance of different areas of systems, it is possible to identify the site with a change. This will determine the place of contact.
These two approaches provide a sufficiently high reliability of recognizing the attacks on the policeman. But that is not all. There is one more component. There are several microphones on the bulletproof vest that constantly listen to external sounds. Special software responds to the sound of the shot. If, in a short period of time, breakdowns of bulletproof vests and a shot of sound were recorded, then it is possible with a fairly large guarantee to say that they are interrelated. Knowing the difference in time, you can determine the approximate distance from which the shot was fired. Having data from several microphones, you can triangulate the location of the arrow. Having a shot record, you can try to determine the make and model of the gun.
There is also the idea of using cameras on the back and chest that take pictures of an attack.
Pistol and holster
Now let's talk about a system that learns that the policeman was shooting. In the gun shop is a small sensor, which says that the store has one cartridge less. If, at the same time, the microphone system says that it has detected the sound of the shot, the policeman has fired a shot, which will be immediately reported to the central post.
Another system monitors the pistol holster. Several sensors report whether the gun is in the holster or not. In addition, they measure the method and speed of pulling the gun. In different situations this happens in a special way. For example, when a policeman pulls a pistol in a dangerous situation or when he pulls it out for surrender to a gun.
To reduce the chance of error, one more component will be used: a special bracelet on the wrist. The bracelet can play the role of a watch strap. The bracelet has a sound sensor, an accelerometer and a compass. At the moment of the shot, the hand twitches that the accelerometer notices. The timing of the sound of the shot must also coincide. If the sensor in the store says that it is one cartridge less, then it can be argued that the policeman fired.
A magnetic compass is located in the store, so you can always look in which direction the shot was fired. By combining these data with the accelerometer readings in the wrist strap, it is possible to accurately reproduce the trajectory of the shot. Naturally, the compass readings in the strap and in the gun magazine must match up to a certain degree.
By combining the data from the holster sensors and the sensors on the wrist, you can determine whether the pistol pulled a policeman, or the pistol was taken by strangers. For example, the holster reported that the gun is removed, and the sensor on the wrist says that the hand is stationary. In this case, you can safely beat the alarm, because It is quite possible that the policeman does not control the situation and has just been disarmed by outsiders.
One of the slides mentions the possibility of "linking" a pistol and a wrestling tracker by determining the distance between them. In case the gun was removed from the holster, but it is too far from the wrist tracker, this is interpreted as not a standard situation. Separately considered the moment when the policeman – right-hander with a tracker on his right hand took the gun in his left hand. According to the logic of this subsystem, this is a non-standard situation and an alarm will be raised, which may be incorrect behavior. However, the very fact that a policeman – a right-hander holding a pistol in his left hand is already suspicious.
The system integrated into the shoes should track gait change. If the policeman suddenly limped, his gait changed dramatically, this is a clear sign that something is wrong with the policeman. In the end, the policeman can get wounded in the leg and all the previously mentioned systems will not notice this.
Judging from the footnote at the bottom of the slide, the system with shoes remained at the idea stage and was not realized. One can only assume that probably at the design stage, the implementation difficulties were discovered.
Non-standard patterns of behavior
At the end of the presentation there is a slide entitled "non-standard behavior patterns". Unfortunately, it has a description of exactly one pattern: running. And really, it's hard to argue, how long have you seen a policeman "calmly running about his business"?
The brain of this system at the moment is an ordinary smartphone. Special software accepts data from all subsystems, searches for known patterns and sends a message to the central server in case of a positive result.
Unfortunately, the presentation does not say anything about the method of data transmission from subsystems to a smartphone. For example, if you use Bluetooth, it is not clear how to solve the problem of quality connection and guarantee the transfer of data, which in this system plays a critical role.
It is also not clear why this project is closed and why it became known only now, after the leak. At first glance, there is nothing top secret in the project.
The system will continue to work reliably if any subsystem fails, and in some cases even if several subsystems fail.
At the moment there are closed discussions on the topic, but is this system necessary for ordinary policemen? The main argument against its introduction is the cost and the rarity of the cases when a policeman gets under fire. In addition, while it is absolutely unclear how long this system will be able to operate on one battery charge, because Some functions require quite a lot of processing power. It will be silly if the accumulators are not enough for even one shift of a policeman (an average of 12 hours).
A very different story, if we are talking about special police units. But in this case, the system should take into account that, perhaps, there is a serious skirmish involving a dozen people. It must correctly interpret all incoming information from all sensors, especially microphones and accelerometers … With the right approach, this system can reliably determine the number and location of shooting criminals.
An important point that was envisaged in the development of this warning system is The fact that it is not necessary to purchase new bullet-proof vests and pistols. In the existing bulletproof vests layers are being introduced. The weapon is not subject to alteration, the sensor in the gun shop does not affect its operation.
The whole system will work completely automatically. In this way, the policeman can focus on the situation and not be distracted by the thought that reinforcement is needed.
The system becomes very logical if you think of it as a standard system from the Internet of Things world. We have a set of sensors that produce raw data. A certain logical system filters them and makes certain conclusions based on known patterns. For example, a holster gives information whether a gun is in it. The next logical level system analyzes data from several sources and uses more complex patterns for analysis and can make more accurate and complex assumptions. For example, analyzing data from a holster, a tracker on a hand, a gun and microphones, allows you to infer whether a shot was fired. The more sensors involved in the system, the more reliable it becomes and the more information it can provide.
It can be assumed that only central ideas are sounded in the leaked presentation. The thought comes to mind, and why not put a pair of sensors in a telescoping baton, in a flashlight or in a gas bottle?