At Amazon's warehouses, autonomous mobile robots drive around moving crates around the facility. As these robots turn on the same paths every day for years, their caster wheel creates a divot in concrete that grows over time. To combat this, Amazon is interested in various ways to install a steel plate over these divots. To choose the right steel plate solution, I was tasked with creating a gantry machine to simulate a robot's caster running over these steel plates fully loaded. After thousand of cycles, Amazon will use the data collected and go with the plate that is showing the least damage.
I designed and built the entire gantry system by following the engineering process. Starting with listing requirements for our timeline, speed, acceleration, load... etc. I brainstormed ideas and researched existing solutions. The design was done in Solidworks. I did some hand calculations to figure out the physics and size the motor and pneumatics.
By the end of my 4-month co-op I was able to order the bill of materials and assemble the machine. I had to overcome a number of problems throughout this journey. For example, supplier delays forced me to quickly pivot to another linear actuator. Amazon also gave a massive change of scope in the middle of the project going from a two-axis to a one-axis gantry which made me have to change a huge portion of my design. In the end, I'm proud that I was able to overcome these challenges to deliver a working machine by the end of my work term. For more detailed information about engineering behind this project, see the report attached below.