00:00:16.279
Hello everyone. Unfortunately, we are not here with K, and Don't See song Kil is not here, but we would like to say hi. We are very happy to be here again and want to say thank you for giving us a chance.
00:00:47.160
The topic of this session is robots, and there are many robots in our lives now. We can buy giant robots on Amazon Japan, and some robots can ride bicycles. This is a Yamaha robot called Motobot. Motobot says there are probably some things in this world that only I can do.
00:01:18.200
This is moto by Yamaha, and it drives a bicycle. The bicycle is not modified.
00:01:36.479
Some robots are smart. Let's introduce ourselves. My name is Yurie Yamane, and my name is Maysayoshi Takahashi. Our cat is our mascot; unfortunately, he cannot travel by air, so he stays at home during this conference.
00:02:05.439
We have developed an MV application and tools for mruby just for fun. We will show you two types of robots. The first one is Lego Mindstorms EV3. In this session, we would like to talk about a self-balancing robot. Do you know what a self-balancing robot is? It's also called an inverted pendulum. A self-balancing robot has mass located above its pivot point, which can make the robot unstable.
00:02:41.080
This robot is a simple demonstration of control theory. In this session, we will make a self-balancing robot using mruby. Do you know about the Segway? The Segway is one kind of self-balancing machine. I had a chance to ride a Segway at Inama University in Hosomi Praim, which is next to Man Prector M. This area is called Ch Ch Andu.
00:03:32.159
At that time, I was a part of Fukuyama University as one of the staff for the Choku Regional Competition of the Ed Technology Robot Contest, officially known as the Ed Technology Robot Contest. The contest aims to develop advanced skills in Ed technology, such as modeling, designing, and developing Ed systems. It is a contest where robots have the same design hardware.
00:04:30.520
In this context, we use two types of robots. The robot we introduce here is called EV3. It has two modes: it uses a G sensor for balance and a light sensor to track a black line on the ground. EV3 traditionally supports programming languages such as C, C++, and Java. From this year, we can use mruby with EV3.
00:05:20.840
We made a mruby environment for EV3, and you can find the code available for research. Unfortunately, we cannot bring the EV3 robot here, so please watch this movie.
00:06:19.240
The movie was taken during the regional competition in the Ch area of the robot contest this year. The red light under the robot is an infrared sensor to track the black line with the light sensor. The robot follows the course and can turn right or left along the line.
00:06:52.440
Okay, you are going to explain how EV3 works. First, we need to introduce its operating system, ECRT. ECRT is one of the AR developed in the topart project. Topas stands for Toyohashi Open Platform for Embedded Real-Time Systems. Toyohashi is a city located in Japan.
00:07:19.319
The name was selected based on the project leader, Professor Takahashi, who is associated with Toyohashi University of Technology. The mascot character of Topas is called Topi. In English, you can read more about the Topas project.
00:08:05.039
The Topas project is widely used in applications; many companies leverage its products. Some examples include synthesizers from Casio, digital pianos, printers by Brother, and hybrid cars by Nissan. These are just a few of the Topas project users.
00:08:42.080
The Topas project has developed various drivers and APIs for devices you wish to connect. For instance, in our case, we already had EV3 connections and drivers. However, while working on this, I found some bugs in ECRT when using mruby, which led me to contribute to Topas.
00:09:33.160
We will be using ECRT's APIs in this presentation. The schedule for the tasks in our program is based on priority management in a real-time operating system. Each task has its own priority. Lower priority tasks do not execute when higher priority tasks are running.
00:10:12.480
As mentioned, we will use three tasks: the most important task is balancing. This task occurs over four microcontroller cycles. We set the balance task as the highest priority. Each task is written in Ruby.
00:10:27.000
The code for balancing reads the brightness of the light from the sensor, adjusting power to the motors accordingly. This process is continuous, allowing the robot to maintain its balance.
00:11:05.000
In addition, we introduce another task to monitor the buttons pressed on the controller. This task is responsible for activating and initiating other essential functions.
00:11:43.600
When designing real-time systems, the critical factor is ensuring that tasks finish within their deadlines. This involves managing tasks based on their priorities, making sure high-priority tasks execute consistently.
00:12:28.080
We utilize alarms in our system, which periodically trigger task executions. These alarms manage the balance task and the button watch task, ensuring that the necessary processes occur even under high-priority task loads.
00:13:02.350
MB EV has this classes; for more information, please check GitHub. Part two: DIY self-balancing robots.
00:13:33.240
EVC is not so popular among the makers. However, can we create our own robots from scratch? It’s not easy, but it’s possible. We faced many challenges during development, yet it is rewarding. Many examples are available online for reference.
00:14:27.760
To build our robots, we need three components: a target board, sensors, and motors. For our project, we used a Raspberry Pi A+, which is cost-effective and works well with smaller batteries.
00:15:17.000
Initially, we started with a Raspberry Pi B; however, due to a hardware failure, we acquired a new A+. We discovered that various microcontrollers might present unique challenges, making code sharing difficult.
00:15:46.719
If you want to create complex tasks, crafting them on the Raspberry Pi can be quite challenging, but we can successfully achieve self-balancing functionalities.
00:16:36.150
To measure the angle of our robot, we require a gyro sensor. We decided to use the LC GD20, a digital gyroscope. While using it, we found that the initial unit did not function properly. Fortunately, the second one worked as expected.
00:17:10.800
We placed the sensor at the center of the robot for stable readings. We also used a Tamiy Motor for propulsion. The actuator controls were wired to the corresponding ports.
00:17:35.480
After setting up the sensors, we utilized a test program to ensure functionality. The motor's speed is controlled through PWM, allowing for variable output based on the required power.
00:18:32.020
As we demonstrated, the robot can move forward and backward depending on the input received from the control board. The main task is to maintain the balance of the robot while executing these movements.
00:19:39.990
In conclusion, we've discussed two robots using mruby. Building robots can be challenging, but it is also incredibly enjoyable. If you are interested, we encourage you to give it a try. Thank you for listening!
