top of page

Prototype Description

Our full prototype consists of multiple assemblies. The main assemblies are the robot arms, Y-axis, aluminum extrusion, and center box assemblies.

 

The two robot arms were modified and optimized from an open source design. It also contains the grippers at the end to grasp the shoelaces. The two degrees of freedom of the arms are powered through two stepper motors while the open/close motion of the gripper is powered through a small servo motor.

 

The Y-axis assembly adds an extra degree of freedom to the robot. The arm is anchored to a platform which moves back and forth across two linear rods. The robot stops when it hits the end-stop at one of the ends. The platform is pulled through a stepper motor and pulley system. Bearings are attached to the platform to reduce friction as much as possible.

 

The aluminum extrusion assembly provides the main support structure for the entire robot. It holds the Y-axis and center box assemblies and allows distances between them to be adjusted to find the optimal reach of the robotic arm to the shoelace. Because of its lightweight quality and adjustable distance, the platform can be made as compact as possible while not weighing heavily. 

 

The center box assembly contains all of the motor wiring to the three PCBs as well as our logic power chip (Teensy 3.6) and power supply. The size of the box is based on the number of components and the relative height between the Y-axis and the top of the box. 

PCB Schematics

The above schematic is for one of the robot arms. There are connections for three stepper motors and three end stop switches. Each motor controls one axis of motion per arm. The end-stops help place an origin or stopping point for each of the axes. Because there are two robot arms, two PCB boards of this configuration are needed. 

This is the main PCB board. It contains the Teensy 3.6 which is the microcontroller our group is using. All the motor and endstop connections from both arm must come to this board because it will control exactly how much each motor will move. The gripper opening and closing is also controlled by this board. Two connections are needed for the gripper. Each gripper will have a sensor also connected to the Teensy. The final three connections on the Teensy are two laser sensors to detect the start and end point of the shoe. The last spot is for the power supply which will provide voltage to all motors.  

CAD Layout and Drawings

Full Shoe Tying Robot Isometric.PNG
Full Platform Drawing.PNG

Prototype Versions 

Version 1 - Winterbreaker

Our first partial prototype was assembled over winter break and thus earned the name winterbreaker. It revolved around an open source robotic arm that could be used to grasp many different types of items. However, it utilized servo motors and it's precision was not enough to meet our team's needs so the robot arm was not utilized at all. On the right is a picture of what it looks like fully assembled.

Version 2

After finding a robotic arm that could reach the full range of motion necessary for tying a shoe, our team found that the rotational degree of freedom offered by the open source arm (MK3) was not needed. It was instead converted into a linear degree of freedom (Y-axis). At first, very small brackets were utilized and the pedestal to hold the shoe was made of wood. The Y-axis was also attached to a wooden base. However, it was able to be moved. The prototype can be seen on the right.The grippers at this stage of the prototypes was the standard version that came with it.

Version 3 - Springbreaker

A lot of changes were made in the next iteration of our robot. Because there was no way to maintain the Y-axis and Center Box completely anchored to a base, a base had to be designed and machined. The material of choice for this base was chosen to be aluminum extrusions because they could be anchored and also translated depending on the size of the shoe used. Thus the aluminum extrusion assembly was created. Two extrusions in the middle also allowed for the new center box to also be anchored securely. The center box itself was laser cut and expanded to decrease the angle of the shoelace with the top of the box. The box was also laser cut so that the power supply and other electronics could fit inside and be anchored inside. The PCB boards containing all the respective wiring were self-made through Eagle. Finally, the brackets of the Y-axis were compacted so that space would not be wasted. This involved concentrating the area of the Y-axis on the brackets instead of utilizing an entire base. To keep the stepper motors hidden, the orientation of the motor shaft went from horizontal to vertical. A preliminary assembly of Springbreaker is shown on the right.

Version 4 - Old Faithful

Our fourth prototype is the one that started achieving the proposed knot motion very consistently. Hence, it was named Old Faithful. Old Faithful addressed some minor issues from Springbreaker. The 8 mm rods used for the Y-axis experienced some torsion due to the weight of both robotic arms. Given that the plastic was also stiffened for our robotic arms, the weight would increase and create more torsion. One of the simple ways to counteract this was to increase the diameter of the linear rods. Our group decided on 12 mm rods and this simultaneously increased the size of the motor brackets. Endstop holders were also added to the back of the robotic arms so that they could calibrate themselves. Old Faithful also featured one of the finalized version of our gripper assembly. Old Faithful is shown in on the right.
 

Winterbreaker.png
Prototype 3.jpeg
Springbreaker.png
IMG_8331.JPG

Final Mechanism

The final iteration of our shoe tying robot featured some minor changes to Old Faithful. The overall width of the Center Box Assembly was increased to help gripper scoop the shoelace more easily. The wiring was also completely fixed through the use of a drag chain. There was concern that the gripper could interfere with a loose wire. Therefore all wires were rerouted through the back of the center assembly in two holes instead of large slots through the side. Finally, the wires were also routed through the middle of the robotic arm links to avoid any gear interference. This was finalized to also add a better aesthetic effect to the overall project. On the right is a photo of our final mechanism.

Final Mechanism Iso.jpg
bottom of page