Sunday, October 29, 2017

Desktop Programming vs. Chromebook Programming

I was so excited to learn that LEGO had created a cloud-based programming environment for EV3 Mindstorms Robots. Every student in our district has a ChromeBook assigned to them in grades 3-12, so I imagined a lot of opportunity for kids to experiment with programming, even when we weren't meeting for robotics.

But not so fast...

Some of the exciting functionality of the desktop programming environment doesn't seem to exist in the ChromeBook version. Data wires, sensor blocks and math blocks are not part of the suite. What this means is the inability for students to experience concepts like amplitude, randomization and exponential changes in motor speed. It will also be difficult (impossible?) for students to conduct experiments since there are no sensor blocks.

So, while it's convenient to have the ability to program on the ChromeBooks in a pinch, we will be sticking to the desktop version for now.



Sunday, October 15, 2017

Welcome to Robotics for the 2017-18 School Year!

The "brain" and "heart" of the LEGO Mindstorms Robot is the EV3 Programmable Brick. It's a combination of a small computer, control center, data manager and power output. The EV3 runs in the Linux programming environment, and has some amazing functions.

One of the first steps at the beginning of the year is to guide students as they explore the landscape of the EV3. There are four main tabs, and under each of these categories is a variety of functions. 

By navigating through the tabs it is possible to view, log and graph data input (like distance, light intensity, changes in position, etc.), control motors manually, save and organize files, connect to computers via bluetooth, and even write simple programs.

The import and output ports on the EV3 brick are differentiated by letters and numbers. Numbered ports are inputs for data from sensors. Lettered ports are outputs for energizing motors. 

You can learn a great deal about the EV3 Programmable Brick from the EV3 Manual. 

As we progress through the year, and the challenges get more complex, it will be important to have a firm understanding of the way this brick works.

Monday, May 15, 2017

The Results Are In!

This was the first year for the maze and line follower competitions, and they were both challenging, engaging, and exciting. The kids dug deep into their creativity and perseverance when presented with these tough tasks. I'm impressed by each and every one of them --those that solved the challenges as well as those who did not.  The final results for each of the competitions are below:

LINE FOLLOWERS:


  1. FLUFFERNUTTER 61.05 SECONDS
  2. LINE DESTROYER 67.28 SECONDS
  3. HI BYE 67.52 SECONDS
  4. MALAPROPISM 70.78 SECONDS
  5. BOB 70.92 SECONDS
  6. EARL GREY 76.16 SECONDS
  7. MEH 89.73 SECONDS
  8. TEST 96.98 SECONDS
  9. ATLAS 183.58 SECONDS
  10. BRICKY 188.89 SECONDS

MAZE SOLVERS:

  1. EARL GREY 36.62 SECONDS
  2. BURRITO 40.52 SECONDS
  3. MALAPROPISM 43.05 SECONDS
  4. DR. DAB-RAH 55.21 SECONDS
  5. CORNY 61.15 SECONDS
  6. MAZESTROYER 79.14 SECONDS





Wednesday, April 26, 2017

2017 RSU 1 Robotics Olympiad, June 1st

For the first time ever, RSU 1 is hosting its own Robotics Olympiad event on Thursday, June 1st in the BMS cafeteria from 6:30 to 8:30. The competition categories* are:


  • Speed Robot
  • Can Clearer
  • Walker
  • Robot Courier
All robots must fit in a 1 foot by 1 foot square. They may not grow in size by unfolding or deploying extensions during their event.


SPEED ROBOT

Robots are electronically timed through a straight course, 18 feet long. The robots must stay on the course to be timed. The fastest time is the winner!

CAN CLEARER

Robots navigate an elevated table to clear 8 empty cans from the surface. The robot to clear all the cans in the least amount of time is the winner. All cans must be cleared and the robot must not fall off the table in order to be scored.

WALKER

A robot that walks with two or more legs must move forward for three feet through a four foot wide course. The robot that walks the three feet in the shortest amount of time is the winner.

COURIER

A robot must carry a small object from the course start to a target at the course end. The robot must not leave the playing field, and after depositing the object in the target, must exit the target boundary. The robot that completes the challenge in the shortest time is the winner.

Mark your calendars for what promises to be another great night of robot competition!

*The events and rules are inspired by Maine Robotics, and the work of Tom Bickford and his staff.

Friday, March 17, 2017

Line Followers and Maze Solvers, Continued

This year is the first time we introduced the Line Following and Maze Solving challenges, and it has been exciting, difficult, rewarding and intense. Both challenges are programming-intensive,  but they still require students to make robots with structural integrity. 


The first part of each challenge is just to solve the problem: follow the black line without deviation from start to finish, and navigate the maze from start to finish. For those who could solve the problem, the next challenge was seeing who could do it the most quickly.



As we have learned, after first creating a functioning robot it's important to make small changes to the physical form and/or the program to improve its overall performance. Taking a systematic approach, making changes in measured increments and recording the variations in performance will lead to the best iteration of a particular  robot. In self-help terms, we want our robots to "become their best selves." In both these challenges the students had to analyze the interplay between varying levels of speed and different degrees of accuracy. They found that the faster the robot goes, the less accurate it is, but the more accurate it is, the slower it goes.





For the line following 'bots, everyone used the light sensor to follow the black line. Some used the sensor to locate the color black, while others used the sensor to measure reflected light intensity (RLI). The maze solving 'bots had a lot more variety in their sensors and programming ideas. Many used an ultrasonic sensor to follow along the side of a wall and a touch sensor to recognize dead ends, while others attempted to solve the maze with dead reckoning or a series of true/false (Boolean) statements.
In the end, it was the interactive robots (not the dead reckoning robots) who solved the maze. There are too many variables in the course and the behavior of the robots is also not exact. These kits are sophisticated, but they are not high-precision scientific instruments.

In an exciting twist to the maze competition at BMS, one robot solved the challenge and did multiple trials with slight changes, getting better and better each time. Another robot solved the maze once, and the team went back to the drawing board, frantically making changes until the last possible minute. Their second attempt was more than 2.5 times faster than their first with these few key changes.

As our middle school LEGO Robotics program evolves, it's exciting to see the inventiveness and ingenuity generated by the students in an atmosphere of friendly competition. Everyone is looking forward to our final challenges at our competition on Thursday, May 25th.

Wednesday, February 1, 2017

On to the Next Challenges

The next projects we will be working on are a line-following robot and a maze-solving robot. Both tasks require the use of sensors. Threshold levels need to be tested, measured and analyzed. We ask ourselves "what is the target percentage of reflected light intensity we are looking for," and "what is the default distance the robot must maintain from the wall?"

Once the threshold levels have been set, we have to decide what our program is going to do. The best way to begin thinking about this is to write a set of commands in pseudocode. This is text written in lines as opposed to building a program with virtual blocks in the LEGO Mindstorms language. For the line follower, let's say the target RLI (reflected light intensity) is 30%, and that we want our robot to follow the right side of the black line. The pseudocode might look like this:


  • Start program
  • enter loop
  • enter switch
  • If  RLI is greater than 30%, turn left
  • If RLI is less than 30%, turn right
  • continue loop forever
This basic program will need to be debugged and tweaked to find the best balance between speed and accuracy. There is also a way for the robot to change its amount of turn based on the reading of the sensor. This is called a proportional program.

I challenged myself to solve the maze problem, and it was much harder than I anticipated. After writing my pseudocode I toiled to come up with the successful program below. It took many rounds of trial and error, and in the end I included specific sounds to keep track of which parts of the program were being executed.
And here's the 'bot in action:




As soon as we get some line-follower robots, I'll post those too. I look forward to lots of lightbulbs going off as the kids learn to write these relatively complex programs!


Friday, January 13, 2017

RSU 1 SUMO CHAMPIONSHIP 2017





We held the third annual SUMO Robot challenge Thursday night, with 18 teams competing.  You can read a nice write up about it in this Times Record Article . For the first time this year, the event took place in the BMS cafeteria which turned out to be an excellent venue. Spectator's seats were arranged in a circle around the Sumo ring, and a pit area was set up off to the side. 



After receiving feedback from spectators last year that they had difficulty seeing the action, we devised a way to project the bouts onto a big screen at the front of the room. I put my iPhone on a tripod and sent the signal to a projector. It was a big improvement that we'll continue to use.









A former 5th grade student of mine, Joshua Doughty, volunteers to help with our robotics students. He is now a junior at Morse High School, and did a fantastic job refereeing the matches.
We used a web application called "challonge" to build our single elimination bracket. Mr. Seymour took the computer data and transferred it to a big white board for everyone to see.


The competitive spirit was pretty intense, but we tried to make sure that everyone shook hands before and after the match. Regardless of what place a robot took, each team worked tirelessly to build and program it.



The winners, Robbie Fitzmaurice and Soren Langord pose in front of the bracket.
The complete results are listed below. Look for more robotics challenges coming up this spring!


Supreme Champion:

ATLAS (BMS, Robbie Fitzmaurice, Soren Langord)
Worthy Adversary:

BOB (BMS, Saydee McCall, Dyllan Arsenault)

3rd place tie:

Amazed D (WCS Jackson Fuller, Gavin Baillargeon)
Dr. Debra (BMS, Riley Dunn, Dalton Johnstone, Eli Torrey)
5th place tie:

Turtle 2 (BMS, Eleanor Carrolton, Nick Gaynor, Matt Doughty)

Fluffy Kitten Destroyer of Worlds (BMS, Cody Stewart, Bjorn Langord)

KadenMarc (WCS Marc Bradford, Kaden Lopez)

Ram (Patten Free Library)
9th place tie:

Bricky (BMS, Aidan Shields, Jonah Smith, Dylan Philbrick)

Failbot (BMS, Coco Cashman, Sophie Booth, Marissa Miles)

RoboSumo (WCS Natalie Emmerson, Jean Paul)

Gremlin (Patten Free Library)

Star Killer (Patten Free Library)

Honeybadger (WCS Oscar Hennin, Alex Maccio, Jacob Behnke)

Meh (BMS, Emily Tetzlaff, Hadley Wong, Miranda Burpee)

Destroyer (Lucas Martin, Alex, O'Camb)

17th place tie:


The Niffler (WCS, Iris Hennin)

Pencou (WCS Riley Gale, Micah Longbottom)