derrick_chi

Hello all.

I've been planning for sometime now to add some custom circuits and programmable logic to my RCs, I've already done so with my car and I was planning to start on a pretty large project with a monster truck I just got. I plan on adding some autonomous capabilities to the truck by means of custom circuits and programming. Eventually I plan on have the truck at almost 100 autonomous, with the RC option still intact of course. Some of the things I plan on adding to start things off are listed below.

1. Collision prevention.
2. Obstacle detection and avoidance.
3. Autonomous gps navigation.
4. Roll over detection and avoidance.
5. Engine overheating protection.


There will be additional things to add, to be sure, but those are a few of the basics. So what I wanted to know from you guys is if any of you would be interested in a build thread like that? I don't want to start the thread if it wouldn't be interesting or fun you guys to follow along and maybe even contribute. So if you would be interested just chime in and so, if not just don't reply at all, and if I get enough replys I'll start the thread. Oh yeah if you have any questions, fire away.

Anomie

I, for one, would be fascinated by your build and would follow it along. One of my RC projects has spilled slightly into the programmable realm, and I have had plans to build in at least a bit of autonomy on the order of obstacle avoidance.

It sounds like an interesting build, and I hope you do post it here [sm=thumbs_up.gif]

rcnitrohead

It would be interesting, but wouldn't most of those features kind of kill the fun in R/C?[&:] I mean the only one of those features that would be handy IMO would be engine overheating protection.. But nonetheless, interesting[8D]

DaveG55

By all means, post a build thread. I'm sure lots of people, myself included, would be intrested.
And I do have a few questions and comments already.
What kind of car? 1:1 or rc?
Numbers 1 and 2 seem like the same thing - or at least extremely similar.
How exactly will a gps nav system work with a vehicle as small as an rc when gps accuracy is not really all that precise
Will your circuts be able to function fast enough and with enough finese to prevent rollovers?
What measures do you plan to use to bring engine temps down if they do increase to unacceptable levels?
How much weight and volume will all the sensors, actuators and circutry add to the vehicle?

derrick_chi

rcnitrohead

You make a good point some of the features could, as you say "kill the fun of r/c" thats kinda why I wanted to get some opinions from some of you all first, and maybe even some suggestions as to some usefull features to add. The autonomous navigation and obstacle avoidance are actually features for another project of mine of which they are absolutly necessary, I just want to use the truck as a test bed for those features. I don't want to bore anyone, thats why I'm asking about interest first, and also input and help in some areas of the project. With that said...

The truck will retain all of its r/c capability, and all features added can be disabled.

derrick_chi

DaveG55

Thanks for responding, and your interest. I'd like to answer a few of your questions to start things off.


1. For this build I'd be using a 1/8 scale monster truck (Redcat's Avalanche, I'll post pics). In fact I have the truck at me local hobby shop now getting some surfaces added to hold the Microprocessor, and gps receiver.

2. You are right, 1 and 2 are similar with a few key differences.
A. Collision prevention( at least what I had in mind) refers to not allowing contact by basically stopping or slowing the car or truck before a collision, kinda like the new cruise control systems on high end cars today.
B. Obstacle detection and avoidance refers to the car or truck detecting an object and taking evasive action maneuvers to avoid contact, with or without decreasing speed.

3. The rollover prevention part might be a little more trickery that I once thought, I designed a circuit before to perform this function and it did work but it was on a touring car, it probably wouldn't work effectively on a truck so my method and design approach will have to change, and any help you guys can provide in the way of knowledge on that issue or advice would be very helpful. The microprocessor(Quik Silva), I'll be using to make all of these functions possible right now currently operates at over 125MHz in a spartan 3 FPGA, and that roughly equates to a little over 100, 000, 000 instructions per second so its plenty fast and should be able to detect and react in time.

4. This is exactly the kind of stuff I wanted help on from you guys. The only thing that came to mind that made sense was adjusting the throttle response if the temp rose above a preset level. I figured bringing down the rpms would help cool the engine or at least stop it from getting hotter, and actually cutting it off if the temperature did not improve. But any suggestions you guys have on that would be great. I was just talking to the guy at the hobby shop about that and he suggested adjusting the carburetor, or fuel intake (I've forgotten now, forgive me if I'm way off).

5. All the circuits I'll need to add will fit in the palm of your hand, all of them together, every thing will be on the micro scale even the gps receiver. I've pasted in some links below so you can check some of them out.

Hope these answers help clear up your questions.

https://www.navtechgps.com/Shop/equi...d_6206-l-t.asp

http://www.acroname.com/robotics/parts/R93-SRF04.html

http://www.acroname.com/robotics/parts/R217-TR-GPM.html

SpikeX

I'd love to see a thread on this project and I would be very interested to see it come along!

I also have a few questions:

1. How would all this circuitry be powered?
2. Do you know where everything would be mounted, and where the wires will be routed?
3. If there is a fault in the system, will it be easy to diagnose and will it impair the functioning of the normal R/C car at all?
4. How much would everything cost?
5. What are you planning to do with this technology after it is all completed and tested?

derrick_chi

Hello Spikex

Here are the answers to your questions.


1. The circuitry I will be using is all micro-scale and will be powered by the receiver which Quik Silva ( my microprocessor ) will interface with to control all of the steering and throttle servos. All of the circuitry I will be using can be powered at 3.3vs or 5v. The truck will never be in danger of having its circuits damaged from the additional circuits I'll add, they just don't pull enough current, and it will retain all of its r/c capabilities.

2. Actually I have the mounting surfaces being installed now at my local hobby shop, I kinda left the details up to them. In all I only need about 5 very small circuits, listed below. The room required for these circuits is minimal.

a.range sensors
b. FPGA
c. gps receiver,
d. temperature sensor
e. this one is open at the moment.

3. I put in some links in my last post up above checkout the links and those prices you see there are the high end cost of those circuits. But I expect to spend between 250 - 300 US. Which is about the cost of a good nitro engine.

4. Very preceptive, most of this technology is for another project, of which I can't discuss at this time, but its going to be just as fun.

Hope these answers help.

SpikeX

So the microprocessor will be connected to the servos? What would happen if the microprocessor were to fail, what would happen then?

Would you ever consider selling the circuitry for R/C Trucks?

derrick_chi

SpikeX

Yes the microprocessor will be connected to the servo and the receiver, and ever other unit in the truck, it is the focal point of the entire embedded control system. The failure rate on an fpga now adays is really low, I've been using them for over 8 years now and I've never had that problem, however if it were to fail the truck would just completely shut off, so I wouldn't expect much damage.

As far as the selling the circuitry, we can always discuss that, sure. In addition if you need any custom digital circuit design for your RCs let me know we'll talk it over and come up with something. I perfer using fpga to implement all of my ASICs in R/Cs, and they are small enough to embed in r/c.

bellp

I am very interested in this topic, where I hope you decide to keep a build thread. I am actually going to buy a 1/8 monster truck and was thinking of designing an embedded circuit for additional features such as autonomous driving, collision detection, logging sensor data, and etc. I would look at accelerometers and piezo gyros. I think they could work well togeather to determine the forces acting upon the car during turns and jumps.

derrick_chi

"I would look at accelerometers and piezo gyros. I think they could work well togeather to determine the forces acting upon the car during turns and jumps."

Good point. I was thinking about adding the gyros for heading control and to prevent drift, not too sure about accelerometers though.

Do yo plan on implementing you design in a FPGA or on a bread board with actual ICs?

nitrorcjunkie

This is probably the most interesting thread I've seen in the MT forums. Reading the posts got me thinking about all the technology available in our 1:1 rides and how little of that tech has made it's way to our hobby, with the exception of electric RC, with LiPo's and the electronic speed controls, BL motors, etc. I'm strictly surface RC, so I'm not speaking of any other facet of RC modeling.Engine monitoring and control is is the logical application of electronics and computer control.

An interesting thought just occurred to me- Cadillac has an active suspension control that uses some very cool technology that includes struts that employ a shock fluid that changes viscosity when an electric current is applied- think of the possibilities in RC shocks and differentials!

SpikeX

nitrorcjunkie, I fell the exact same way. As a matter of fact, I'm launching a new section of my website soon called "The Future Of R/C" which showcases all the new technology that is being made by people for R/C cars, and what new technology people would like to see on future generations of our little R/Cs. I have loads of ideas of my own, but I don't want to get off topic here.

derrick_chi

I am really glad to see that you guys feel the same way I do about the technology thats out there which we don't see in R/Cs.

Snaut Rocket

I am really excited about this new build. Cant wait till its done.
It seems like you done your home work on this one.
I just hope Money isnt your primary limiting factor.

derrick_chi

Money shouldn't be an issue on this build the items aren't "really" that expensive the most expensive part of the project is the truck, lol, everything else after that is less than half that price. I gotta warn you guys this project will take time there will be a lot of programming and asic design, so once it starts be patient. It promises to be fun though, and I'll try and give as many pics, videos and anyother info you'll request along the way.

SpikeX

So how's the project coming on? Anything new?

tmaxx_man87

it all sounds good except it will break easy, expencive,extra wieght,doesnt make it faster or stronger... however it would make a truck not break as often due to no collisions...im intrested...whared you go?

derrick_chi

Hi guys

I'am still in the hardware building phase. I basically have the Microprocessor designed and built, I am in the process of building the Bus that will serve as an interface between the microprocessor and the sensors, pwms, receiver, and the gps unit. I am almost finished with the bus design as well, it should only take me a couple more days. I just finished designing the I2C Master controller (for the range sensors and the temperature sensor), the RS232 unit (to talk with GPS receiver) is done as well, and so are the PWMs and PWMEs ( Pulse Width Modulation Encoders ) I'll be using to control the throttle and steering and also to receive and encode signals from the receiver/transmitter for control when in manual mode. Right now I'm just putting it all together and that wil take a couple of days.

Once I have all the hardware stuff done, then I just need to write programs for the different functions I want to implement in the truck, thats when I'll start the custom build post. I'll probably need some tips from the experts on certain things when I start to write programs to implement the functions I mentioned before.

I think I want to start off with a "Break in Function" which automates the break in process. This shouldn't be too hard to do but I want to make sure I'm doing everthing right by putying in the proper safety features to ensure nothing happens to the engine. I should begin programing by Monday.

As far as the weight added, really all of the hardware I'll be putting in the truck can fit in the palm of your hand and will weight less than half a role of dimes, remember most of the hardware will be inside FPGA, the rest is sensors and such, all of the stuff is micro in size and so its all very lite weight. Shouldn't be too easy to break but that is a possibility. The total cost of all the hardware will probably be around the cost of an upgraded nitro engine about 200 U.S.

snoking16

if you can engineer an automated break in for a somewhat low cost you could mass produce it and sell it to all the newbs who dont want to screw their engine up.

derrick_chi

I would probably consider that if I can make the process generic enough so as it would be applicable to a variety of different cars and trucks. And I think that is possible, shouldn't be too hard.

derrick_chi

OK so I been doing a lot of searching over the last couple of weeks and what I found is the following. I am going to have to build my own temperature sensor from a few parts the Max6675 (Thermocouple to Digital Converter) and a basic thermocouple. However thats only half the battle because the max6675 uses motorola's 3 wire serial protocol called spi, so I am going to have to design a master controller for the spi bus so as to allow for communication between quik silva ( my microprocessor ) and the sensor.

Since I had to go through this kind of trouble for a circuit who's parts only cost around 15 bucks total, when I get the SPI master controller done and the circuit working I'am going to give the design away free. That ought to disturb some people.

derrick_chi

F.Y.I For those of you who are still waiting on the projects to take off, here is the latest.

At present I am testing the hardware, and things are going well, I only have a couple more tests to run and I'll be able to start the first build thread.

Its taking a while because I had to design and test all of the interface I/O modules which include, Pulse Width Modulators (to facilitate communication between the processor (Quik Silva), receiver and the servos), Encoders, I2C Master controller ( to communicate with the range sensors), SPI Master (to communicate with the temperature sensor, which i had to build as well), a Master Local Bus controller (to facilitate communication between the microprocessor and all of the IO modules), and a BCD ( Binary Coded Decimal) circuit to communicate with the seven segement display ( which will be used to display the values of temperature, object distance, and a few more informational things.


Of course during the first round of testing I am going to uncover bugs of all kinds and I'll have to fix those but that seems to be going pretty smooth, so far, so with any luck I should be done with in the next week or so.

Yub, yub, cmdr!

I am currently involved in a similar project, but with 1/10 electric touring car. I am using an ATmega128 RISC-based microprocessor powered directly by the LiPo battery to execute a SLAM algorithm and try to beat human racers.