First, you check the Bluetooth module connection by upload the simple RC car control code with the android app. If the system is ok, then you upload your EEG code. Here, some questions will arrive?

1. You used hardware serial or software serial?

2. You take digital or analog data to control the car?

3. You used the voltage divider method to connect Arduino with Bluetooth module RX pin or not?

Here, I attached a .ino file and an android app to check your arrangement.

https://play.google.com/store/apps/details?id=braulio.calle.bluetoothRCcontroller&hl=en&gl=US

Make sure the Bluetooth module connected as Md Anowar Hossain mentioned and drivers also then try to use this code https://github.com/Simonsimu/Mind-Wave-Control-Robot/blob/master/SourceCode.pde

I would like to add another solution following Mohammed Tawfik . If you are able to acquire a Raspberry Pi, I suggest using a Pi to control the robot car over MQTT from your PC.

In this scenario, your PC via some sort of software (lets assume you have a program that can translate EEG readings into useable data) parses it into a command signal (defined by you, think about a basic ladder logic code). Your PC can very easily "publish" this code to a MQTT broker to which the Raspberry Pi will "subscribe" too and translate it into motor driver command (for Raspberry Pi there should be quite a few motor driver shields available).

The big advantage is that MQTT is far more easier to work with than Bluetooth and is a lot faster than it. This is because the transmission speed is as fast as the wi-fi router that will act as the wireless communication hub between your PC and the Raspberry Pi.

While my proposed idea may seem daunting at the beginning you actually don't need any special hardware to pull this off. The raspberry pi can itself act as both broker and client. Here 'client' is analogous to the code running in Arduino. Additionally, if you have some Python experience this won't take long. Furthermore, you won't have to fiddle with any wiring to external devices such as the HC05, which speaking from personal experience is not a good idea to be used with on a breadboard. When I built the experimental setup for this paper https://ieeexplore.ieee.org/abstract/document/8934569, I used HC05 which often lost connection when the robot was moving around.

My concluding thoughts would be, if you attempt to go this proposed, the first challenge would be to get the data from the EEG device to a python script in your PC first. Once you are able to do that, the rest would not be as difficulty provided you are able to get hands on the correct hardware.

Good luck

Md Anowar Hossain

I really thank you so much and i will start working on this pattern .

I really dont know the difference between hardware serial and software serial am just using the USB to compile and upload the code .

2- I take digital data to move the robot .

3- I am really sorry , am trying to learn what is the voltage divider.

i will start working on your advices , million thanks .

Mohammed Tawfik thank you so much for the code , so what you mean that in Arduino I write the blutooth code that Md Anowar Hossain gave me and then on the same file I write the your car moving code right !?

Azmyin Md Kamal your Idea is amazing , as matter of fact I managed to transformed the raw EEG data into usable data saved on Excel file and did some analysis on weka by using Random Forest Algorithm and I managed to get over 90% , so I am really excited to replace the arduino for more efficient data for smoother robot .

But can the Jetson nano replace the raspberry Pi because I already have it !?

many thanks all of you and I will do my best and show you the results .

Abdulrahman Alsulaimani Jetson Nano maybe an "overkill" but yes it is a far better replacement. I myself am using a Xavier NX for my thesis research and it is quite powerful. However, the jetson nano won't last as long as the Pi owing to Jetson nano drawing more power than Pi. Since you have jetson nano, I strongly suggest going with Deep learning based method than Random forest since you would need to do the inference in real-time. If I am not mistaken, Scikit learn (Python's ML module) runs on the CPU.

I suggest using pytorch v1.7 and cuda-toolkit v10.2.89 since the jetpack img (a modified ubuntu 18.04) comes pre installed with CUDA 10.2 and OpenCV 4.4.1. Using nvidia's GPU to do deep learning calculation will drastically speed up the inference speed and allow you to control the robot car even better. What's even better is Pytorch is just Python with OPP so it is much easier to learn than TensorFlow (this is my personal opinion).

We can discuss this more over zoom if you prefer. I would love to learn more about this project and if possible go for a collaborative work.

Good luck,

Azmyin

Azmyin Md Kamal I would love to this is my discord name , please add me so we can arrange the meeting .

Abdul#2702

I really look forward to the zoom meeting , also I have a question .

as for arduino uno i bought the esp8266 wifi bridge so i can transmit the data from my sensor to the pc and then to the car but I did all the codes and adding the library for ESP8266 and it wont upload , even if it does upload how people use it afterwards....!?

any idea guys Mohammed Tawfik Md Anowar Hossain

You can easily transfer data from your sensor to PC or server. There are some free server available, thingspeak, adafruit dashboard e.t.c. If you want, you can transfer your data to this server via esp8266 module and remotely monitoring your data.

I think, this report book helps to you.

https://www.researchgate.net/publication/336774803_Remote_Sensing_IoT_based_Android_Controlled_Robot

Abdulrahman Alsulaimani I would prefer Zoom over Discord. I will send you a zoom invite over private message sometime in the weekend.

From the issue you mentioned, it seems you have overcomplicated things. My reason for suggesting Jetson nano is it supports MQTT protocol with just using two python libraries which are very simple and straightforward to use. With a wi-fi router in between, communicating between the nano and the PC would be a simple matter. I dabbeled with ESPs before but did not find its coding easy to approach. But then again, it was quite a while back and I haven't had a chance to see the latest developments.

Md Anowar Hossain I have the sensor, PC , and the esp 8266 and when i used the code to upload it to arduino it says ( Faild to connect to esp8266 ; Timed out waiting for packet header ) .

PS : I've download the library of the esp8266 .

I just wanna connect the esp8266 to a network so I can send my data , my data is on emotive BCI OSC i can broadcast it but the only problem now is the connection of esp 8266 .

Azmyin Md Kamal no problem I just said the discored so you can send me the link but no problem tell me the day , time and which country do you live in so I can check out the link for meeting .

As for the nano I really liked the idea and i am gonna work with it after the connection is good to go .

am really grateful for knowing you guys .

Do you complete your board setup for esp8266?

https://github.com/esp8266/Arduino

Azmyin Md Kamal I know I overcomplicated things sometimes and it only happens when am close at something .

I did the almost everything but I think I need the esp8266 to stream my data from the sensor to arduino and I really got lost in arduino codes and arduino example .

sorry guys for asking so much but I really appreciate the help and many thanks .

Md Anowar Hossain it came to my mind does the hardware shape of esp 8266 matter !?

Well ... here is mine ..

I did everything just like the link you gave me and really thank you but still i get the same error ....

Is the problem from the examples in arduino !?

I also tried the ( think speed , adafruit dashboard ) but i got stuck not knowing how to function it .

I also tried the platform IO and it wont compile right .....

Abdulrahman Alsulaimani don't get discouraged. The process of learning is making a lot of mistakes.

Try to post your problem with arduino and esp in Stack Overflow. Often you will find people who had already faced this same issue and solved it.

Try these following steps

1. Does your EEG sensor output in 3.3V or 5Vs? As you may be aware of the fact that esp like MCUs are usually not 5V tolerant.

2. Think of the problem as small chunks rather than one large problem. Write them down if that helps. I have often debugged complicated problems in my codes using this method.

3. I would strongly suggest trying out the MQTT route. It is far easier to implement and is quite straight forward.

Good luck,

Azmyin

I think this links might be helps to you.

https://forum.arduino.cc/index.php?topic=622116.0

https://github.com/espressif/esptool/issues/490

https://stackoverflow.com/questions/65173101/esp8266-esptool-fatalerror-failed-to-connect-to-esp8266-timed-out-waiting-for

https://randomnerdtutorials.com/solved-failed-to-connect-to-esp32-timed-out-waiting-for-packet-header/

Azmyin Md Kamal I might complain a lot but giving up is not in my dictionary I know most of the research requires time and sometimes even more than we thought but the feeling you have when you finish your research is Priceless .

I will keep going with your solutions and i will keep you guys updated if you don't mind .