Imagine a problem - two people disappeared in the forest. One of them is still mobile, the other lies in place and cannot move. The point where they were last seen is known. The search radius around it is 10 kilometers. It turns out an area of 314 km2. You have ten hours to search with the latest technology.
The first time I heard the condition, I thought, "Pfft, hold my beer." But then I saw how advanced solutions stumble over everything that is possible and impossible to take into account. how about 20 engineering teams tried to solve the problem dozens of times easier, but did it to the limit, and only four teams did it. The forest turned out to be a territory of hidden tricks, where modern technologies are powerless.
Then it was only the semi-finals of the Odyssey contest, organized by the Sistema charity foundation, whose goal was to figure out how to modernize the search for missing people in the wild. In early October, the final was held in the Vologda region. Four teams faced the same challenge. I went to the place to watch one of the competition days. And this time I went with the thought that the task was unsolvable. But I did not expect to see "True Detective" for fans of DIY electronics.
Snow fell early this year, but if you live in Moscow and wake up late, you might not see it. That which does not melt by itself will be scattered by the workers a hundred percent. It is worth driving from Moscow for seven hours by train and a couple more hours by car - and you will see that winter has actually begun a long time ago.
The final was held in the Syamzhensky district near Vologda. Near the forest and a village of three and a half houses, the organizers of the Odyssey set up a field headquarters - large white tents with heat guns inside. Three teams have already conducted searches in previous days. Nobody spoke about the results, they were under the NDA. But the expressions on their faces gave the impression that no one had coped.
While the last team was preparing for the test, the rest of the participants put their equipment on the street for beautiful shots of local television, showed and explained how it works. The Nakhodka team from Yakutia rattled beacons so that journalists who were interviewing had to pause.
They took the test the day before and got the worst weather. Snow and gusty winds did not even allow the drone to launch. Many lighthouses could not be placed because transport broke down. And when one of the devices finally worked, it turned out that a tree had been knocked down by the wind, and it pressed down the button. However, the team is watched with curiosity because they are the most experienced searchers.
- My whole team is hunters. They have been waiting for the first snow for a long time. They will see the traces of any animal, they will definitely catch up with it. I had to restrain them as guard dogs,” says Nikolai Nakhodkin.
Scouring the forest on foot, they would probably be able to find a trace of a person, but such a victory would not be credited to them - this is a technology contest. Therefore, they relied only on their sonic beacons with a powerful penetrating sound.
A truly unique device. It is evident that it was made by people with a lot of experience. Technically, it is very simple - it is an ordinary pneumatic wah with a LoRaWAN module and a MESH network deployed on it. It can be heard for one and a half kilometers in the forest. Many others do not have this effect, although the volume level is approximately the same for everyone. But the right frequency and configuration give such results. I personally recorded a sound at a distance of about 1200 meters with a very good understanding that this is really the sound of a signal.
They look like the most non-technological, and at the same time they have the simplest, most reliable and very effective solution, let's say, but with their own limitations. We cannot take and use these devices to find a person who is unconscious, that is, these products are applicable only in a very narrow range of situations.
- Nikita Kalinovsky, technical expert of the competition
On our day, the last of the four teams worked - MMS Rescue. These are ordinary guys, programmers, engineers, electronics engineers who have never searched before.
Their idea was this - with the help of several aircraft-type drones, scatter a hundred or two small sound beacons over the forest. They connect into one network, where each unit is a radio signal repeater, and begin to make a loud sound. A lost person must hear him, find him, press a button and thus transmit a signal about his whereabouts.
Drones are taking pictures at this time. The autumn forest is almost transparent during the day, so the team hoped to detect a lying person in the photo. At the base, they had a trained neural network through which they ran all the pictures.
In the semi-finals of MMS Rescue, beacons were scattered by ordinary quadrocopters - this was enough for four square kilometers. To cover 314 km2, you need an army of copters and probably a few launch points. Therefore, in the final, they teamed up with another team that had previously dropped out of the competition, and used their Albatros aircraft.
The start of the search was scheduled for 10 am. Before him in the camp was a terrible fuss. Journalists and guests walked around, participants dragged equipment for technical verification. Their tactic of planting beacons in the forest ceased to seem like an exaggeration when they brought and unloaded all the beacons - almost five hundred of them.
- At the heart of each is an arduino, oddly enough. Our programmer Boris made an amazing program that controls all attachments, says Maxim, a member of MMS Rescue, “We have LoRa, a board of our own design with attachments, mosfets, stabilizers, a GPS module, a rechargeable battery and a 12 V siren.
Each lighthouse costs about 3 thousand, despite the fact that the guys had every ruble in their account. There were only two months for development and production. For most team members, the MMS Rescue project is not their main focus. Therefore, they returned from work and until late at night were engaged in preparations. When the parts arrived, they manually assembled and soldered all the equipment themselves. But the technical expert of the competition was not impressed:
— I like their solution the least. I have big doubts that they will then collect the three hundred lighthouses that they brought here. Or rather, how - we will force them to collect, but not the fact that it will turn out. The search itself will most likely work if sowed in such a quantity, but I did not like the drop configuration, nor the configuration of the beacons themselves.
- Beacon technology reduces the number of kilometers traveled by feet. The beacons that will be scattered now suggest a further hike through the forest to collect. And it will be a distance that does not reduce the amount of human labor. That is, the technology itself is okay, but perhaps we need to think out tactics on how to scatter, so that later it would be easier to collect, says Georgy Sergeev from Liza Alert.
Two hundred meters from the camp, the drone team set up a launch pad. Five planes. Each takes off with a slingshot, carries four beacons on board, scatters them in about 15 minutes, returns and lands with a parachute.
Lost hunters
After the start of the search, the camp began to empty. The journalists dispersed, the organizers scattered around the tents. I decided to stay all day and see how the team would work. Some of the participants were still engaged in the control of drones, some got into the car and drove through the forest to place beacons along the roads manually. Maxim stayed at the camp to watch the net unfold and receive signals from the beacons. He told me more about this project.
“Now we are watching how the network of lighthouses is unfolding, we see the beacons that appeared on the network, what happened to them when we saw them for the first time, and what is happening now, we see their coordinates. The table is filled with data.
Are we sitting and waiting for a signal?
— Roughly speaking, yes. We just have never scattered 300 beacons. Therefore, I look at how you can use the data from them.
- On what basis do you scatter them?
“We have a program that analyzes the terrain and calculates where to drop beacons. She has her own set of rules - here she looks into the forest and sees the path. First, she will offer to throw off the beacons along it, and then she will go into the forest, because the deeper, the less likely that a person is there. This is a practice that was voiced by rescue teams and people who got lost. I recently read that a missing boy was found 800 meters from his house. 800 meters is not 10 km.
Therefore, we first search as close as possible to the likely entry zone. If a person got there, then he is most likely still there. If not, then we will incrementally expand the search boundary. The system simply grows around the likely point of human presence.
This tactic turned out to be the opposite of that used by experienced search engines from Nakhodka. On the contrary, they calculated the maximum distance that a person could walk from the entry point, placed beacons around the perimeter, and then closed the ring, reducing the search radius. At the same time, the beacons were arranged so that a person cannot go beyond the ring without hearing them.
— What did you develop specifically for the finale?
“We have changed a lot. We did a lot of tests, measured different antennas in the forest, measured the distance of the signal. In previous tests, we had three beacons. We carried them on foot and attached them to tree trunks at a short distance. Now the hull is adapted for dropping from a drone.
It falls from a height of 80-100 meters at a speed of 80-100 km / h of drone flight, plus wind. Initially, we planned to make the shape of the hull in the form of a cylinder with a wing sticking up. We wanted to place the center of gravity in the form of batteries in the lower part of the hull, and the antenna would automatically rise up in order to achieve good communication between beacons in forest conditions.
"But they didn't?"
- Yes, because the wing into which we inserted the antenna interfered greatly with the aircraft. Therefore, we came to the form of a brick. Plus, they tried to solve the issue of nutrition, because each element is heavy, it is necessary to cram the minimum mass into a small case while maintaining the maximum amount of energy so that the lighthouse does not die in an hour.
Improved the software. 300 beacons in the same network can interrupt each other, so we did the spacing. There is a big complex task.
It is necessary that our sirens at 12 V scream as they should, so that the system lives for at least 10 hours, so that the arduino does not reboot when LoRa is turned on, so that there is no interference from the tweeter, because there is a step-up device that gives 40 V out of 12.
- And what to do with a lying person?
“Unfortunately, no one has given a reliable answer to this question. It would seem wiser to search with dogs by smell along the fallen trees. But it turned out that dogs find far fewer people. If the lost one lies somewhere on a windbreak, theoretically it can be photographed and recognized from a drone. We have two planes flying with such a system, we collect data in the air and analyze it at the base.
How will you analyze the photos? See everything with your eyes?
— No, we have a trained neural network.
- On what?
— On the data that we ourselves collected.
When the semi-final was over, experts said that in order to find people using photo analysis, a lot of work still needs to be done. The ideal option is when the drone analyzes real-time images on board using a neural network trained on a huge amount of data. In reality, teams had to spend a lot of time uploading footage to a computer, and even more time reviewing, because no one really had a working solution then.
- Now neural networks are used in some places, and they are deployed both on personal computers, on Nvidia Jetson boards, and on the aircraft themselves. But all this is so raw, so undereducated, says Nikita Kalinovsky, - as practice has shown, the use of linear algorithms in these conditions worked much more efficiently than neural networks. That is, the definition of a person by a spot on the image from a thermal imager using linear algorithms in the shape of an object gave a much greater effect. The neural network found almost nothing.
- Because there was nothing to learn?
- They claimed that they taught, but the results were extremely controversial. Not even controversial - there were almost none. There is a suspicion that they were either taught incorrectly or taught the wrong thing. If you correctly apply neural networks in these conditions, then most likely they will give a good result, but you need to understand the entire search methodology.
- We recently launched , says Grigory Sergeev, “While I was here at the competition, this thing found a person in the Kaluga region. That is, here is the real application of modern technologies, it is really useful for search. But it is very important to have a carrier that flies for a long time and allows you not to smear photos, especially at dawn and dusk, when there is practically no light in the forest, but you can see something else. If optics allows, this is a very good story. In addition, everyone is experimenting with thermal imagers. In principle, the trend is right and the idea is right - the question of price is always worrying.
Three days before, on the first day of the finals, the search was conducted by the Vershina team, perhaps the most technologically advanced of the finalists. If everyone relied on sound beacons, this team's main weapon was the thermal imager. Finding a market model that can produce at least some result, fine-tuning and tweaking it - all this was a separate adventure. In the end, something happened, and I heard enthusiastic whispers, as a beaver and several elk were found in the forest with a thermal imager.
I really liked the decision of this team precisely by ideology - the guys are looking for technical means without involving ground forces. They had a thermal imager plus a three-color camera. They searched only with letaks, and at the same time they found people. I will not say whether they found the one they needed or not, but they found both people and animals. We compared the coordinates of the object on the thermal imager and the object on the three-color camera, and determined what it was exactly from two images.
I have questions about the implementation - the synchronization of the thermal imager and the camera was done carelessly. Ideally, the system would work if it had a stereo pair: one monochrome camera, one three-color camera, a thermal imager, and everyone works in a single time system. Here it was not. The camera worked in one system, the thermal imager in a separate one, and they had artifacts because of this. And if the speed of the letak was a little higher, this would already give very strong distortions.
- Nikita Kalinovsky, technical expert of the competition
Grigory Sergeev spoke most categorically about thermal imagers. When I asked his opinion on this in the summer, he said that thermal imagers were just a fantasy, and in ten years the search party had never found anyone with their help.
“Today I see the price drop and the appearance of Chinese models. But while it's still wildly expensive, dropping such a thing is twice as painful as the drone itself. A thermal imager that can show something decent costs more than 600 thousand. The second mavic costs about 120. Moreover, a drone can already show something, but specific conditions are needed for a thermal imager. If for one thermal imager we can buy six mavics without a thermal imager, naturally we will operate with mavics. It is not worth fantasizing that we will find someone under the crowns - we will not find anyone, the crowns are not transparent to the greenhouse.
While we were discussing all this, there was not much activity in the camp. Drones took off and landed, somewhere in the distance the forest was overgrown with beacons, but no signals were received from them, although half the allotted time had already passed.
At the sixth hour, I noticed that the guys began to actively talk on the radios, Maxim sat down at the computer, very anxious and serious. I tried not to go in with questions, but after a few minutes he came up to me himself, quietly swearing. A signal came from the beacons. But not from one, but from several at once. After a while, the SOS signal was trumpeted by more than half of the units.
In such a situation, I would have thought that these are problems with the software - the same mechanical malfunction cannot come out at the same time on so many devices.
We ran the tests two hundred times. There were no problems. It can't be software.
A few hours later, the database was overflowing with false signals and a bunch of unnecessary data. If at least one of the beacons worked on pressing, Max had no idea how to determine it. Nevertheless, he sat down and began to manually sort through everything that came from the devices.
Theoretically, a real lost person could find the lighthouse, take it with them and move on. Then, perhaps, the guys would have detected the movement on one of the units. How will an extra portraying a lost person behave? Will he also take it or go to the base without a device?
At about six o'clock, the guys who were working on the drone ran to the headquarters. They uploaded the photos and found very clear human footprints on one of them.
The tracks ran in a thin strip between the trees and disappeared outside the photograph. The guys looked at the coordinates, compared the photo with the map and saw that it was on the very edge of their flight zone. The tracks go north, where the drone did not fly. The photo was taken over five hours ago. Someone on the radio asked what time it was. He was answered: "Now is the time of our flight."
Max continued to dig into the database and found that all the beacons began to beep at the same time. They had something like delayed activation. To prevent the button from working during the flight and fall, it was deactivated for the duration of the delivery. That is, the lighthouse was supposed to come to life and start making sounds half an hour after departure. But along with the activation, the SOS signal also worked for everyone.
The guys took out several beacons that they did not have time to send, opened them up, started sorting through all the electronics, trying to find what could go wrong. And a lot could go wrong. When the electronics were tested, they were not yet packaged in a case that could withstand a reset. The solution was found quite late, so several hundred lighthouses were assembled by hand at the last moment.
Max at this time manually sorted out all the messages from the beacons in the database. One hour remained until the end of the search.
Everyone was nervous, me too. Finally, Max came out of the tent and said:
- Write there in your article so that you never forget to screen.
Having dismantled several lighthouses, the guys were hooked on the theory. Since the housing for beacons appeared very late, all the electronics had to be packed more compactly than planned. And due to the fact that time was running out, the guys did not have time to shield the wires.
A few minutes later, a signal was found in the database from a device that worked much later than the others. This lighthouse was delivered to the forest not by drone, the guys brought it themselves and tied it to a tree next to one of the roads. The signal came from him at half past one, and now it was already half past seven on the clock. If the button was indeed pressed by an extra, then due to the noise, the signal from it could not be recognized for several hours.
Nevertheless, the guys perked up, quickly wrote down the coordinates of the beacon and the time of activation, and immediately ran to fix the find.
Much was at stake, and the technical experts reacted to the find with disbelief. How, among a bunch of broken beacons, could there be one that really worked? The boys tried hard to explain.
Let's take a step back. Did changing the case cause your signals to stop working after a fall?
- Not certainly in that way.
- Is it related to the body?
- This is due to the fact that the SOS button worked before the moment it should work.
- Did it activate when it fell?
- Not when falling, but when a sound signal is triggered. The beep sounded peak-peak, 12 V was converted to 40 V, the wire was wired, and our controller thought that the button was pressed. This is still speculation, but very likely to be true.
- Very strange. She cannot give such leads. I highly doubt it. The reason for false positives in terms of circuitry?
I'll explain now, it's simple. Previously, the case was wider and the distance between the elements was greater. At the moment, some wires, including the wire from the button, run right next to this thing.
- Is it a transformer?
- Yes. And not only with him. It raises by 40 V, this is a booster. There is also a 1 watt antenna nearby. When transmitting, we receive a certain message, and immediately it goes into the SOS state.
- How is your button tied to a percent?
- They just hung it on the GPIO, with a pull-up on the bottom.
- You hung the button directly on the port, pulled it down and any signal that passes through it immediately jumps up, right?
- Well, it turns out so.
“Then it looks like the truth.
- I also already realized that it was necessary to pull the wrong way.
Have you tried wrapping the wires with foil?
— Tried. We have several such lighthouses.
- Well, you saw that when the signals go on the beep, and when the signal comes through the antenna, you have ...
- Not certainly in that way. Not when there is a signal on the beep, but when it is time to activate the beacon. The button is cut off so that it does not accidentally press on a branch or something else when flying in an airplane. There is a certain time delay. When the time comes to turn on, activate the button, the entire beacon turns on, as if they had taken and closed the power to it. No delays, nothing, all the elements immediately started to rise and work, and at that moment the button is triggered.
Why doesn't everyone work like that?
Because there is an error.
- Then the next question. How many products worked on false positives? More than a half?
- More.
- How did you single out one of them, which was submitted as the coordinates of the missing person?
“Our captain went by car to the most likely areas and carried the beacons by hand. He took a box that contained a separate batch of beacons, and actually arranged those beacons that did not have such an error. We analyzed the data we collected, isolated from everyone who did not start shouting SOS at the time when it should be activated and went to the lighthouse, which started shouting SOS much later than 30 minutes.
- Do you admit that there was no false positive at first, and then it could appear?
— Well, you know, it has stood for more than 70 minutes since the lighthouse was revived. We analyzed the coordinates - this is not far from the place where, according to legend, a man appeared.
Half an hour before the end of the search, the team finally received the coordinates of the missing person. It looked like a real miracle. There is a mountain of lighthouses in the forest, more than half of them are broken. Worse, half of the beacons from the batch that were placed by hand also broke. And in an area of 314 square kilometers strewn with broken beacons, an extra found a worker.
It just needed to be checked. But the team went to celebrate a possible victory, and I, after eleven hours in the cold, could leave the camp with peace of mind.
On October 21, about a week after the test, I received a press release.
According to the results of the final tests of the Odyssey project, aimed at developing technologies for the effective search for missing people in the forest, the integrated system of radio beacons and unmanned aerial vehicles of the Stratonauts team was recognized as the best technological solution. All developments presented in the final were finalized with the funds of the AFK Sistema grant fund in the amount of 30 million rubles.
In addition to the Stratonauts, two more teams were recognized as promising - Nakhodka from Yakutia and Vershina with their thermal imager. “Until the spring of 2020, the teams, together with rescue teams, will continue to test their technical solutions by participating in search operations in the Moscow and Leningrad regions and Yakutia. This will allow them to refine their solutions for specific search tasks,” the organizers write.
MMS Rescue was not mentioned in the press release. The coordinates that they transmitted turned out to be incorrect - the extra did not find this beacon, and did not press anything. Still, it was another false positive. And since the idea of a continuous sowing of the forest did not find a response from the experts, it was abandoned.
But the Stratonauts also failed to cope with the task in the final. They were the best in the semi-finals too. Then, on an area of 4 square kilometers, the team found a person in just 45 minutes. Nevertheless, experts recognized their technological complex as the best.
Perhaps because their solution is the golden mean between all the others. This is a balloon for communication, drones for surveying, sound beacons and a system that monitors all search engines and all elements in real time in real time. And at least this system can be taken and equipped with real search parties.
“Search today is still the Stone Age with rare flashes of something new,” says Georgy Sergeev, “Unless we go with ordinary torches, but with LED ones. We are not yet at the stage where the men from Boston Dynamics are walking through the woods, and we are smoking on the edge and waiting for them to bring us the missing grandmother. But if you do not move in this direction, do not move the whole scientific thought, nothing will happen. We need to stir up the community - we need thinking people.
Source: habr.com