Chassis

To physically separate mechanical and electrical components, a hardboard box design was chosen for the chassis. With the exception of the drivetrain system, all mechanical components were mounted on the removable top panel, and all of the electronics were mounted on the bottom panel, thus completely enclosing them and protecting them from potential shorts caused by the water present in the challenge.

Cutout holes on the bottom panel allowed access to buttons, switches, and a USB port on the TINAH board, which allowed us to upload updated control code onto the robot very quickly. Hole cutouts were made along the side walls of the chassis to allow cables to be cable-tied to, as well as to reduce overall weight.

All mechanical and electrical components were mounted onto the chassis using non-permanent fasteners, providing modularity and allowing components to be removed, redesigned, and remounted without affecting the rest of the robot. Hardboard was chosen as the material of choice, as it is electrically insulative, can be fabricated very quickly, and the additional strength and rigidity offered by sheet metal was not required.

Arm Mechanism

The 3 degrees of freedom agent collection arm was designed to provide maximum robustness and flexibility. Using two large servo motors for positioning and a small servo motor for gripping, the coordinates for each arm traversal command could be tuned easily using software.

Because of the parallel sponge design on the gripper, the arm was able to grab bath toys without requiring a large amount of positional accuracy, and it very rarely released the toys unintentionally once they were held. The four bar linkage mechanism for the gripper allowed precise tuning of design parameters, such as sponge angle and overall gripper dimensions before and after gripping.

Due to different requirements on each joint, each servo motor was mounted using different methods, using gear ratios whenever the full 180 degree rotation was not necessary on the output movement. The optimum mounting method for each motor was developed through many iterations of trial and error, and the resulting motor mounts involved minimal backlash while allowing easy replacement of motors due to non-permanent fasteners and friction mounts.

Collection Box

The collection box, made of hardboard, was designed primarily to have minimal weight while remaining structurally secure. For that purpose, holes were cut out of each face, just small enough so the bath toys are unable to slip out.

Attached to the box is a rubber band loaded acrylic hook which is initially held down via a retractor mechanism at the start of each run and released only when the elevator raises the platform supporting the collection box. This both lowers the initial height, allowing the robot to pass through the first gate, and prevents interference with the claw during the rescuing agents phase.

Given the low mass of the box and bath toys, we opted for a thin rounded hook to reduce rotation about the axis of the hook's arm which was significantly reducing the speed of the box's descent down the zipline. The hook was designed such that the centre of gravity of the entire collection box aligned vertically with the centre of the hook between two bearings for the even loading and optimal smoothness during descent.

Elevator

For the elevator, we initially leaned towards using a timing belt and pulley system. However, after prototyping it was quickly discovered that using a closed loop of belt meant that the platform would not be able to descend past the lower pulley or even ascend all the way to the upper pulley due the physical constraints applied by the mounting of the platform to the timing belt. Due to this, we pivoted towards using a pulley and string system, a change that only required a few adjustments to the design.

To drive the elevator, we used a small 224:1 geared DC motor, which had the torque necessary to lift the platform, yet could be powered and controlled directly from the TINAH microcontroller board due to its low current requirements. As the motor's shaft was rectangular, it could easily be directly mounted to a 3D-printed pulley. A drawer slide was attached to each side of the platform to constrain the platform's movement and orientation in all but the intended vertical direction.

TINAH

Microcontroller Board

reneZ uses a TINAH board as its microcontroller. The TINAH is similar to an arduino, but was created for the ENPH 253 class to provide more functionality for our robot. Some of the features include:

• 4 DC Motor outputs
• 3 Servo outputs
• 2 buttons and 2 knobs for on board control
• LCD

H Bridge

Motor Control Circuit

The motor outputs on the Tinah board can only provide a DC motor 9V and 1A. We used Geared Barber Colman motors which needs ~12V and 1.5A to operate. In order to get enough power to these motors we needed to make a separate H-bridge. A H-bridge is used in order to control the speed and direction of a DC motor, essential for driving our robot.

IR Detector

Signal Processing Circuit

During the competition our robot needed to detect infrared beams of light at two frequencies, 10Khz and 1Khz. In order to achieve this we used a signal processing circuit which would take the corresponding frequency and amplify it, while attenuating other frequencies.

Class Overview

This is how we structured our robot to run the competition course. It is structured from high level classes to low level environment classes. The controller is the main class of the robot. It is in charge of handling the finite state machine and communication in between components. The driver controls the drive train. It uses QRD sensors to follow on black tape using PID correction. IR detector is used for reading infrared signals. The claw class is used to rotate and position the claw using 3 servo motors.

Competition Flowchart

This is a flowchart overview of the tasks needed to be accomplished during the competition.

Finite State machine

We saw using a finite state machine was the easiest and most robust way to accomplish all of the tasks in the competition, instead of using some complex system. We used a finite state machine to model each component of the competition in our robot. We broke the competition up into 4 parts:

1. Driving to the gate: This state accounted for driving up to the gate and detecting when the gate is on/off
2. Driving to agents: During this state, we needed to adjust speed and PID correction to get the fastest speed up the hill.
3. Agent collection: This state is for going around the circle and collecting each of the agents using the claw.
4. Agent dropoff: This state is for driving using infrared sensors (instead of tape), raising the collection box, and delivering the agents down the zipline.