Electromechanical Robot Design

Overview

I built this robot as part of MIT's Design and Manufacturing I class (2.007). The objective is to build a remote-controlled, electromechanical robot that scores the most points by completing tasks on a competition board in the allotted time. My robot, "Dobby", placed in the top 20 out of over 100.

Contributions

I designed and built this robot independently, with some guidance from the course staff. My contributions included:

Brainstorming and sketching concepts for the robot
Calculating the required design specifications
Designing components through hand sketches and CAD
Fabricating the robot in the machine shop

Skills

design | Solidworks | engineering drawings | machine tools (e.g., mill, lathe, shear, brake, bandsaw)

Motivation

The primary purpose of the robot is to score points by completing tasks on a competition board:

The robot must also meet certain design constraints, such as starting volume, weight, number of motors, materials, voltage, and total stored energy.

"Mini-Me" Robot

Before designing my robot for the competition, I built a smaller, simple robot to gain familiarity with the tools and design principles I would use for my final robot. The objective of this "mini-me" robot was simply to collect a ball from the competition board and deposit it into a tube.

I started by creating a driveable base, which I designed as a sheet metal part in Solidworks and fabricated using a sheet metal shear and brake, a bandsaw, and a drill press.

I brainstormed concepts for the ball manipulator and prototyped an attachment on the robot base. After testing this prototype, I designed and fabricated the mechanism from sheet metal for the final robot.

Cardboard prototype

Solidworks model

Sheet metal version

Working demo (2x speed)

Parallel Gripper

For my competition robot, I calculated the potential point values for each task on the competition and concluded that the "mindblowers" (pictured right) had the highest value. Points are earned by removing the test tubes from the shelf, with the point value increasing for higher shelves.

I sketched several concept ideas and decided to make my most critical module (MCM) a parallel gripper mechanism with an extendable arm to reach the higher levels. I calculated the force (and corresponding motor torque) needed to remove the test tubes, sketched a plan for the design of the mechanism based on similar examples I researched, and created a 3d model in Solidworks to visualize the kinematics of the system.

I started by fabricating the base parallel gripper mechanism. I modified my initial design based on the stock material available in lab, laid out my general design plan, then created engineering drawings to machine the aluminum components.

Because alignment was critical for this mechanism, especially the gearbox, I drilled the holes on a mill. I cut the stock using a bandsaw since those dimensions were less critical. To securely connect the gears to the lead screws, I designed a connecting piece, which I turned on the lathe.

Engineering drawings for parallel gripper base parts (machined from 1"x3" hollow aluminum box stock)

I prototyped the arm extensions for the parallel gripper. I started with simple foam core arms taped to the gripper to demonstrate that the mechanism had sufficient grip force to grab and remove a test tube. I tested this successfully (with manual positioning since I had not yet built the driveable component). I also started prototyping longer foam core arms that could fold to fit into the starting volume and extend to reach the highest shelf of test tubes.

Parallel gripper with short foam core arms (2x speed)

Foam core prototype for folding/extending arm

I started testing with sheet metal prototype arms, but these were not very effective. The arms were not rigid enough and too slippery, as demonstrated below:

Sheet metal prototype arms - insufficient friction between arms and test tubes (4x speed)

Added tape to arms to increase grip - able to remove test tube but not consistent or repeatable (4x speed)

I tried adding tape and rubber to increase the friction and introducing a bend in the sheet metal arms to increase rigidity, but there was still too much compliance and not enough grip force.

Because the arms were so long and there was compliance at the joints, even though the arms themselves were rigid, there was major deflection and loss of force transmission between the gripper base and the ends of the arms, which I did not account for in my idealized calculations.

I worked on the extension mechanism in parallel, though in hindsight I probably should have just focused on getting the regular arms to work first. The arms could extend but could not grip the test tubes.

Arms in folded (left) and extended (right) position. Arms are held in folded position and extended by elastic bands when activated.

Crank Mechanism and Final Design

When I couldn't get the arms working by the last week, I shifted my focus to a crank mechanism to operate the competition board's elevator. I decided that this would be more feasible to secure points than trying to fix the arms, which I had already been working on for weeks with little success.

I measured the force to turn the crank with a force gauge, calculated the torque, selected a motor and gear ratio, and designed an attachment for the robot to interface with the elevator crank.

I was able to build this mechanism in a few days and successfully use it for the competition, achieving a score in the top 20 out of over 100 participants.

Crank mechanism to operate elevator on competition board. 3:1 gear ratio to increase torque output.

Clip from final competition (4x speed). Robot successfully operates elevator crank (skipping due to broken gear tooth).

Reflections

I initially chose a more complicated mechanism for the "cool factor" and learning experience, so I don't regret trying it even though it didn't end up working, but this did demonstrate the value of simple yet effective designs.

I found that having a clear plan made the fabrication process much more efficient. Once I started running into problems, I didn't have a clear plan for how to address the issues, so I hit a block in the design process. For future projects, I hope to not only have an initial plan but also a better strategy for systematically analyzing and addressing unexpected issues that may arise.

I also learned the value of prototypes and the dangers of relying too much on over-idealized models.

Perhaps most importantly, I found that projects involving physical components take much longer than expected, so it is critical to plan ahead and build in time for unforeseen challenges and iterations.

Regardless of results, this project helped me develop my design and fabrication skills. I learned how to use Solidworks for CAD and engineering drawings and gained more familiarity with tools such as the sheet metal shear and brake, mill, lathe, bandsaw, and drill press. This project turned out a little rough and unrefined as it was my first design class, but I am actively developing my knowledge of best design practices and hope to improve with each project.

S.PARK

in

Contact: