wiki:Other/Summer/2022/Hardware

Context Navigation

Mini Smart Car Hardware Design

Miniature Smart Car Hardware Design

WINLAB Summer Internship 2022

Group Members: Brandon Cheng, Michael Mogilevsky, Aamay Puntambekar

Project Objective

Design and prototype an approximately 1/15 scale smart car platform to be used for experiments in the orbit smart city environment. This project will involve:

3d modeling for the car chassis
Prototyping the electrical systems with the eventual goal of designing a printed circuit board
Programming microcontrollers to work with sensors

This project focuses on creating a platform that can be used for repeatable experiments. Beyond just the construction of the car, the first priority is to create a robust system with sensors to allow for precise calibration. The finished platform will have rear-wheel drive and use a servo for Ackermann steering, which is the type of steering found in cars. Generation of control signals for the motor controller and servo will be performed by a microcontroller such as an arduino. The microcontroller will also be used to ingest data from any sensors on the car such as an inertial measurement unit (IMU) or encoders used for speed sensing. The car will carry an Intel realsense depth camera and UP board (x86 platform with raspberry pi form factor).

Background Material

This group will need to use a 3D modeling software such as Fusion 360, which is available for free to students. Fusion 360 will be used to design the chassis for the vehicle to accommodate the controllers and sensors mounted to it, as well as some additional hardware components. If group members have limited prior experience with 3D modeling software, they should familiarize themselves with Fusion 360. These videos are a good place to start. A computer will be available at Winlab for students to use for 3d modeling.

The parts list for the car is not finalized yet, but some design specifications are known:

Track width (distance between tires on each axle) should be approximately 6 in.
Car will be rear-wheel drive, with a single motor (as opposed to a direct drive motor for each wheel).
Steering will be implemented using a servo motor with an Ackermann linkage, as with an RC car.
The car will need a rear differential to avoid tire slip on turns.

Some of the design choices (like the use of Ackermann steering and the use of a rear differential) are similar to the design of conventional RC cars. However, this project is different from RC cars in several ways. Since we're not prioritizing speed or off-roading, we don't need a suspension or torque dampening. Also, because we're only planning to drive the cars indoors at relatively slow speeds, we should be able to use 3d printed parts in the drive train and get a reasonable lifetime out of them. Projects such as this or this can be used as inspiration or a starting point. We will be able to order additional bits of hardware such as bearings, screws and RC car parts after initial designs are done.

The car will have an Intel UP board as the onboard controller as well as a realsense camera to be used for sensing and remote control. The car will also have a Teensy microcontroller (arduino alternative) which will be connected over serial to the UP board and used for real-time control of the car and sensor readings. The Teensy will be used to send signals to the motor controller as well as read from any sensors on the car. The motor should have an encoder attached to it to read the speed of the motor directly, and we may also want to attach encoders to the wheels of the car to better estimate any slip. The car will also have an IMU with accelerometer and gyroscope.