I'm building a flying drone. Who can help choose materials and give useful advice?

Question

Electronics

I'm going to build a flying drone. Need your help in this matter.
I need to find out what to make a drone from, what components and parts are needed for it.
It is advisable not to buy ready-made expensive components, but to make the material yourself.
As for the currency, it is desirable to do this at no extra cost. because the drone must be made ready for cheap repairs in case of damage and generally made strong enough, both for impacts and for climates with extremes of high and low temperatures.
Ideally, you need to assemble a drone, with the ability to control it without visual contact with the owner and the drone itself (through any device, laptop or android), and you also need to automate it in the sense that if communication with the owner is lost, so that you can switch to autonomous control and go to a safe zone, or to the place where you came from. You also need the ability to send a drone from point A to point B without the participation of a remote control. It is desirable to make the distance at which you can control the drone as much as possible, in such a case, few people can help me, but if you can help, I don’t know how I can thank you or what I will thank you for, maybe with currency .. not the point. If anyone can help let me know.

Answer 1

You here www.rcdesign.ru/. Look for FPV quadcopters
Too much to paint.
As an option, look at dvj phantom as a template drone for your tasks

Answer 2

Start by learning the mechanics of multicopters, these are the easiest drones to build and fly. Make a small quadcopter with Arduino. To start without FPV. If you have experience in programming in JAVA, the design can be simplified by writing a control for Android, then you will not need expensive transceiver equipment, all control from a tablet or smartphone. Such a thing will not fly far, but without FPV you won’t fly far anyway. Later, you can attach a Wi-Fi camera.
When you figure out what's what, you can make an FPV device /
Chassis can be made from anything : plywood, plexiglass, even plastic pipes

Answer 3

          Oum...uff.. You know, there is such a "company" - Geoscan (Geoscan). Here they have all this implemented as it should, even the way you imagine it and a little more. Especially the part about weather conditions and endurance. The price of the issue is about 800k, there is a variant of full support with software and additional. whistles - 1.3 million. I mean that only money will help you in this matter, and the quality of assistance will be directly proportional to the amount, because the project draws on such a medium-sized project to develop a special-purpose UAV.
          Understand that you can only design and assemble at home with a supporting frame. Home conditions are different, but I, for example, will not undertake to design and assemble, say, an engine, even as a worker, although the conditions allow. Not because it's difficult, but because it's expensive. More expensive than buying ready-made. (check out hobbyking, by the way. Quite a contender for the purchase of components).
          So, I tell you right away that you will have to buy everything: engines, a controller, a radio module, an inertial navigation module, GPS, propellers, a battery, a power supply unit, a camera, and everything that is simply unrealistic to make yourself better than it is in the store.
          Therefore, in fact .. I suggest that you first make a list of what you are physically unable to craftfrom raw materials. And this is almost everything except the body. Then pick up these components for yourself on the same king or any other online hobby hardware store and compare with the cost of a ready-made solution (having ready-made software). Why hobby? Because the price tag for professional pieces of iron will definitely not please you. There, one engine can cost under 50k wooden ones. In no case do I urge to score on development, on the contrary, I welcome, but I bring it up that it will not work out cheaply and cheerfully with such a request. Immediately imply that as soon as you save somewhere by buying a node cheaper (i.e. heavier, slower, more gluttonous or otherwise) than the recommended one, you immediately lose twice as much of the desired parameters. Losses then multiply each other even more, yes ..
          Also, in advance, while you were just thinking about this idea and turned here (obviously, nothing has been done in this direction yet), check out the market for ready-made devices. Directly make a sign for yourself - what is on the market, what parameters it has, how much it costs. Read reviews and discussions. Firstly, it will show you how much it really costs in the end. It will be very significant when there will be devices for 7, 30 and 500 kilo rubles before your eyes and that each of them can at least from the one declared by the manufacturer.
          Seriously, in practice, I have never seen what you described even in the hands of people who are perfectly understanding in all the subtleties. Yes, assembling an indoor drone out of cardboard/plywood is a very real task. Play within a room, maybe two, chase a cat with FPV, without getting up from your chair to check the readiness of dumplings. But weather resistance and high autonomy are a completely different class of devices, and autonomy is a different class of software.
          Further, respectively, about software . Well, everything is simple here .. You will have to write a lot and difficult. And the first thing you can already start reading is SLAM ( Simultaneous Localization and Mapping ) and INS ( Inertial Navigation System ). This is where Google is your best friend. The topic is chewed, discussed on the network, articles on Habré aboutINS (search) and SLAM (search) are in abundance. There is nothing special to advise here - you need to study the mathematical part, pick up an IDE and write code. Well, or order its development for a specific hardware, having certain sensors and device parameters on hand. If you don’t understand anything about this yet, then you can buy a simple Arduino and train on it. The functionality that you described will not fit in three lines. We need layouts, tests and configurations. It is elementary to link with the software for managing from the phone ( This, for example ).
          Учитывая вышесказанное и вообще, могу сказать вам, что не имея вообще никакого опыта и представления о БПЛА, вы замахнулись на непосильную задачу. Надежды юношей питают, понятно дело, но такой амбициозный проект с нуля с большой вероятностью обречен на забивание через уже пару месяцев попыток вникнуть. Так что я бы советовал не мучить себя и окружающих, купить средней пакости квадрокоптер, тысяч за 60 и пристроить для своих задач. Fixid привел отличный вариант - DJI Phantom. Это 100% будет дешевле и быстрее при прочих равных.