Final Project¶
Final Project (tentative)¶
Basics¶
A modular and maze with electronic features to be used in robotics training tutorials and competitions.
Problem description¶
I work in the Robomania section at our National Science Centre. We run robotics demonstrations, workshops, camps and clubs. One of the best kits for teaching robotics is the Lego Mindstorm EV3 kit. One of our most useful challenges is getting students to build and program a robot to solve a maze and pass through it. The current maze we use has some major problems:
-
The maze itself is bulky and heavy, making it difficult to transport when we run the activity at remote locations
-
The maze design is fixed. We give students free range to use whatever programming techniques they want to “solve” the maze. Rather than come up with an intelligent program that can solve a variety of maze configurations, some students use a brute force technique (e.g. programming the robot to go 10cm forward, then turn left 90°, then 30cm forward, turn 90° right......). This requires a lot less thought about the robots algorithm.
-
Difficulty is fixed because the maze is fixed. When we encounter particularly talented students who come up with a working solution quickly, we can’t adjust the mazes diffculty to give them a harder challenge.
-
The maze is a little plain. It’s part of a robotics challenge, so having just a wooden maze with no shiny techy features isn’t as visually enticing as it could be.
Solution¶
A Modular maze that can be:
-
Easily broken down and set back up for easier transport.
-
Reconfigured into various layouts to add more challenges.
-
Has electronic features to make it more exciting. (electonic timer and barrier).
Solution (first draft)¶
-
Maze base will be designed like a peg board with uniform holes thoughout. They will be made of individual pieces (likely plywood) about 1.5’ x 1.5’ that can be joined together to make the larger whole base.
-
Maze walls pieces will be made of plywood or MDF with appropriately spaced pegs on their base to fit into the holes on the base. The wall pieces will be made to the same height, but varying lengths so that they can be configured in a variety of layouts.
-
Grooved brackets will be cut to reinforce the walls to add more stability to the chosen layout. Wall pieces may need to have grooves cut into their ends to accomodate brackets.
-
Grooved paths along wall pieces for clean running of electrical wires for sensors/servos. (potential future upgrade)
-
Electronic control. Conected to proximity sensors (or presure plates) that control the timer and barrier gate. Servo for barrier gate. Display for timer.
-
3D printed mounting brackets for sensors and servo(s). 3D printed electronic box for display and circuitry.
Solution (2nd draft)¶
Maze parts:
-
I have decided to abondon the peg board design on the floor panels. I did a brief test with MDF sheets and encountered the following problems. In order for the pegs from the wall pieces to have enough stength so that they didn’t break or crack when a robot hit the walls, the pegs had to be a little “fatter” than initially intended. This lead to the following problems:
-
The peg holes in the floor were too big. Most of the robot designs used a “ball and castor” rear “wheel”. This ball and castor is a standard part in EV3 kits and is included in allot of sample robot designs. The problem is that the ball tended to get caught in the peg holes, and caused the robots to go off course. I don’t want to add any additional constraints to the students robot, so this caused a problem.
-
The thickness of the peg holes meant that I had to build the walls out of thicker material. This makes them heavier and makes the material more expensive.
-
-
Instead I plan to use flat square floor panels (probably 0.5m x 0.5m each) with small holes in each corner. The holes will allow me to create small 3D printed brackets that will link the floor panels together and hold them in place. These brackets will also come flush with the surface of the floor panels so there shouldn’t interfere with any robot’s mobility.
-
The maze wall pieces will also now be greatly simplified. They will just be flat rectangular panels, all with the same height, but varying lengths so that they can be combined into any maze design we like. Not having to align with floor pegs means that we can easily place wall pieces at any angles to one another if we want to make a rather tricky maze layout.
-
To hold the wall pieces in place, I plan to design 3D printed brackets that will grip onto the outside edges of the floor panels and the outer walls. aLl the inner walls will be connected together with a variety of 3D printed brackets on top that will hold them to one another.
Electronics:
-
Because of the mocular design of the maze, the electronic component is also modular. The control board is a development board with pins to easily connect to any breakout boards for input and output.
-
The maze competition is a competition to complete the maze in the fastest time. So the main output from the board will be a screen that will output the robots time. A POTENTIAL future upgrade would be to also add a servo or step motor to control a “traffic arm”, the raising of wich will signal the start of the timer.
-
The maze timer will be stopped by the robot crossing the exit of the maze. This will be detected by an ultrasonic distance sensor that will detect when an object passes in front of it.
-
The modular nature of the maze and of the development board means it will be easy to change or add features to the elctronic component of the maze without having to make adjustments to the main maze body. Potential future upgrades way be:
- Adding a raised traffic arm and/or sound buzzer to signal the start of the timer.
- A pressure plate or additional ultrasonic sensor to let the robot start the timer itself when it enters the maze.
- A pressure plate or IR sensor to indicate when the robot has left the maze and stop the timer.