Thursday, June 21, 2012

Workshop Days 3 and 4 (June 20-21)

 For the last two days we have been mainly working on the project for the workshop that deals with planning the path of a robot. Each team has been assigned a name based on the name of a robot - STAN (Stanley - Stanford U.'s autonomous car), SFOX (Silver Fox UAV), TNAO (Team Nao - the humanoid robot). Each team has to plan the path of the robot so that it can print its name using the Sharpie pen attached to it. This project illustrates many fundamental aspects of the motion of the differential wheels-based e-puck robot such as how long and at what speed to rotate its wheels to control its motion, waypoint-based navigation, etc. Teams which finished working on this project started working on a robot dance - flashing lights and moving the robots in a rhythm. We will demo these projects at the end of workshop presentations on Friday. Pictures and videos from the two days are below.




epuck pen pattern












Wednesday, June 20, 2012

My Problem and Solution

Hot days and months that offer little to no rain can impact lawns and gardens everywhere. For years we have used sprinklers to water our lawn to give it water when the weather is dry. Since the weather can be so unpredictable, your lawn might need a lot of water or none at all. What if you are on vacation and the weather all of sudden becomes very dry? A sprinkler system that senses heat can fix this problem without the need to re-program your sprinkler to water the lawn. 

Hygrometers and heat sensors (sensors) would detect the temperature and the humidity to determine whether or not to water the lawn for an hour or not at all. The sprinkler heads (actuators) would then water for longer periods of time or none at all. The hotter and drier the weather is, the more it gets watered. If the weather is humid and rainy, the lawn will get watered for a short time or not at all.

A sprinkler system can cost 2-4 thousand dollars. The new sprinkler system will need to have a heat sensor and hygrometer modification, which would cost about 50 dollars.  Labor cost won't cost much if the sprinkler system is being modified, which is about 5 dollars.
My problem is all the wasted space and material by handicap ramps and parking space. My idea for a solution is to use a larger version of the modular robots in a snake instead of wheels on a wheelchair. It would eliminate the need for ramps and the problems with curbs and stairs/escalators. The sensors of this object would be the person driving it, and the actuators would be the ModBots powering it in a snake. It could be a nonprofit organization and since these are a new invention, I don't know the price, but the only labor is bolting them on and building a chair.

Tuesday, June 19, 2012

Workshop Day 2 (June 19)


Today we looked more closely at how to move robots while avoiding obstacles in their path. We started with an introduction to a model of the e-puck robot inside the Webots robot simulator. We looked at a program that built on yesterday's mybot program to avoid obstacles, but runs on the e-puck robot instead of the mybot.

We then started exploring on the dynamics of the e-puck robot by writing a small program called speed_test. We specifically looked at how to rotate the two wheels of the robot at different speeds gives rise to different trajectories for the robot such as a straight line, a curved path, small and large circles, and finally, spinning the robot on its axis. After completing the program on Webots, we switched on the e-puck robots and connected them via Bluetooth to the computer running Webots. We then transferred the speed_test program to the e-puck robots.


We discovered that the simulated robots run much faster on the simulator than the physical robots. We adjusted the speed of the robots and tried different trajectories with them.

Then, we worked on using the pen on the simulated e-puck robot to make the robot draw patterns on the ground inside the simulator. We will use this program in tomorrow's (Wed) class by attaching Sharpie pens to the physical robots to make trails and draw patterns on paper.

 
In the latter part of the class, we wrote a program on how to work with the LEDs on the e-puck robot. The e-puck robot has a ring of 8 red LEDs around it. It also has an orange colored front LED and a green colored body LED. We wrote a small program to turn one or two LEDs on and off in a blinking fashion. Then we worked on a more complex program that does a basic robot 'dance' - it blinks the LEDs around the robot one after the other, and finally blinks the front and body LEDs, while moving the robot forwards, backwards and sideways.

.Once again, we worked with the program first on the Webots simulator and then transferred it to the physical robot. We started exploring how to control the speed of the robot and blinking frequency of the LEDs to make more 'dance' patterns.


For the last 15 minutes of today's session, the students were given a tour of the C-MANTIC Lab to show the larger robots, landmine detecting robots and a simulated modular robot called ModRED (along with a video of the physical ModRED robot). 

Tomorrow, we will continue work on the robot dance and start working on making patterns on paper with e-puck robots.

Monday, June 18, 2012

Workshop Day 1 (June 18)

Day 1 (June 18)

Pre-workshop: The first day of the robotics workshop. The workshop will be held in PKI 383. There are going to be 6 participants in the workshop.

Post-workshop: We ended up with 7 participants. Parker, Preston, Connor, Jack, Casey, Alex and Drew. In the first part of the class we went through introductions, had a short quiz of robots in movies, looked at pictures and videos of some real robots. We briefly looked at the e-puck robot and some of its basic functionalities like following a wall and avoiding obstacles. 









In the latter part of the class, we started working with the Webots robot simulator and programmed a simple simulated robot called mybot to avoid obstacles using its two front IR (infra-red) sensors. 






The students were also given a homework to identify an important real-life problem that could be automated using robots, and suggest the materials and costs of the robot(s) that would be appropriate for solving the problem. Due on Wednesday at 11 PM. Answers can be posted on this blog site. Students are also required to have discussions on each others ideas on the blog site.