Published June 9, 2020 © GPL3+

First Steps with M.A.R.K

M.A.R.K is a robot designed for learning about Artificial Intelligence. Easy to assemble and use, I had a chance to test him.

BeginnerProtip2 hours470

Things used in this project

Hardware components

seeed M.A.R.K

Software apps and online services

Codecraft from Tinkergen

IDE for graphical coding that generates Python code for several products, such as M.A.R.K

Story

Introduction

Make A Robot Kit, or M.A.R.K. for short, is a robot designed to make artificial intelligence and robotics easy to learn and accessible to all, that should be available in July (it is currently on KickStarter ). The MCU used for its intelligence is a K210, with which I had a already worked before, so when I got offered a chance to try MARK before it was even released, I couldn't say no.

First: the assembly

MARK comes with everything needed to build it and get it running, and it could hardly be easier

MARK in his box

After attaching the board to the chassis, the sensors are basically plug-and-play: attach them to the chassis, plug them were they should (everything is written on the board) and you're good to go !

the board

To make it look better than a bunch of cable on wheels, a magnetic chassis comes on top of it to give MARK this look:

MARK, ready to run

Second: the code

MARK can be coded in Python, or using a graphical IDE. On my first project with a K210, I had used Python, so I decided to give the graphical IDE a try.

Once again, it was really easy to get started with it: all usual functions as well as MARK specific functions are available, and you can check in real time the Python code generated:

View of the IDE

Since MARK is made for AI and machine vision, I decided to make him simply look for a face and go toward it.

Several models are preloaded on MARK, and the functions to use them are available within the IDE, so it was quick to achieve:

First, as long as no face is detected by the camera, MARK turns on himself

When a face is detected, MARK stops (plays with its lights to show that he saw it) and, depending where the face is seen, starts moving again to be right in front of it:

Once the detected face is centered in front of the robot, it once again lights its lights to show it is ready, and then starts going toward the the face, until it's at 5 cm of it:

And here it is, a robot who will come toward you, and will even bring you the salt when you're too lazy to get up to get it from the other side of the table if someone gives it to him first

Conclusion:

This is a really small and easy project that can be done to get familiar with MARK and its environment, before moving on to more complicated applications using even more of his modules (he comes with a line follower, an electromagnet, a speaker, some buttons, and some packs are available with an accelerometer, a light sensor, a rubber band slinghot, or a grabber)

I will try to post some more advanced projects with MARK soon !

Python code generated by the graphical IDE

# generated by Codecraft for cyberEye
# codes make you happy
from gpio import *
from modules import ws2812
import utime
from maix_motor import Maix_motor
import time
from camera import *

ws2812_13 = ws2812(fm.board_info.D[13],5,2,3)

Face = FaceDetection()

def start_handler_0():
    pass
    ws2812_13.set_led(1, (0, 255, 255))
    ws2812_13.display()
    ws2812_13.set_led(0, (0, 255, 255))
    ws2812_13.display()
    utime.sleep_ms(250);
    ws2812_13.set_led(1, (255, 0, 0))
    ws2812_13.display()
    ws2812_13.set_led(0, (255, 0, 0))
    ws2812_13.display()
    utime.sleep_ms(250);
    ws2812_13.set_led(1, (0, 0, 0))
    ws2812_13.display()
    ws2812_13.set_led(0, (0, 0, 0))
    ws2812_13.display()
    Maix_motor.servo_angle(2, 135)
    while True:
        time.sleep(0.02)
        if not Face.getFaceDetectionResult():
            Maix_motor.motor_motion(1, 3, 0)
        else:
            Maix_motor.motor_run(0, 0, 0)
            ws2812_13.set_led(1, (0, 255, 255))
            ws2812_13.display()
            ws2812_13.set_led(0, (0, 255, 255))
            ws2812_13.display()
            utime.sleep_ms(500);
            ws2812_13.set_led(1, (0, 0, 0))
            ws2812_13.display()
            ws2812_13.set_led(0, (0, 0, 0))
            ws2812_13.display()
            utime.sleep_ms(100);
            if Face.getFaceDetectionProperty(1,'x') > 100:
                ws2812_13.set_led(0,(0, 255, 255))
                ws2812_13.display()
                Maix_motor.motor_motion(1, 3, 0)
            else:
                ws2812_13.set_led(1,(0, 255, 255))
                ws2812_13.display()
                Maix_motor.motor_motion(1, 4, 0)
            while not (Face.getFaceDetectionProperty(1,'x') < 90 or Face.getFaceDetectionProperty(1,'x') > 110):
                time.sleep(0.02)
                lcd_display_string_row(Face.getFaceDetectionProperty(1,'x'), 1)
            Maix_motor.motor_run(0, 0, 0)
            ws2812_13.set_led(1, (0, 255, 255))
            ws2812_13.display()
            ws2812_13.set_led(0, (0, 255, 255))
            ws2812_13.display()
            utime.sleep_ms(500);
            Maix_motor.motor_motion(1, 1, 0)
            while not get_ultrasonic_distance(4) < 5:
                time.sleep(0.02)
            definition_fun()

def definition_fun():
    pass
    Maix_motor.motor_run(0, 0, 0)
    ws2812_13.set_led(1, (0, 0, 0))
    ws2812_13.display()
    ws2812_13.set_led(0, (0, 0, 0))
    ws2812_13.display()
    while True:
        time.sleep(0.02)

lcd.clear()
start_handler_0()

Credits

Bapt G

2 projects • 0 followers

Embed the widget on your own site

First Steps with M.A.R.K

First Steps with M.A.R.K

Things used in this project

Hardware components

Software apps and online services

Story

Introduction

First: the assembly

Second: the code

Conclusion:

Code

Python code generated by the graphical IDE

Credits

Bapt G

Comments

Related channels and tags