MOONYOUNG'S RESEARCH
  • Home
  • Work Experience
    • KAIST Hubo Lab
    • iRobot
    • inTouch Technologies Inc.
  • Projects
    • Embedded Computer Vision for Robot Localization
    • Vision-based Robotic Arm
    • CUAUV
    • Exoskeleton Arm
    • Robotic Segway
    • Robotic Arm Manipulation
    • Multi-core Processor
    • Piezosensor on Bipedal Robot
  • Blog

iRobot Corporation

Picture
Picture
iRobot is a pioneering robotics company based in Bedford, Massachusetts that is known for Roomba vacuum cleaning robots. In addition to the 15 million home-product robots sold worldwide, iRobot developed Unmanned Ground Vehicle for the military and remote telecommunication platforms for businesses. 

Role: Systems Engineering Lead Intern
Duration: Jan 2016 - June 2016
Tasks: 
  • Completed an end-to-end design of the Velocity Validation System to support performance claim on all Roomba models.
  • Developed fixture and software to enhance claims testing adherence of a controlled test lab to the IEC standard
  • Performed testing and analysis for robot's mobility and cleaning performance 

Velocity Validation System

All Roombas are rigorously tested in a controlled lab according to the IEC 62929 standards at iRobot. To validate the velocity at which Roomba claims are made, I used photoelectric infrared sensors and Raspberry Pi to capture velocity. This was an end-to-end project involving CAD, PCB design, and user-interface scripting.

PCB design

Used Altium Design software to create 2 layer PCB that handle analog sensor input and handle power distribution. 

CAD and Installation

Used Creo software to CAD sensor mounts for the break-beam to slide between Robotunit frames​. 

User Interface Software

Wrote a batch script to SSH into RaspberryPi, launch the Python script, and transfer the data back. The Python script handled user interface and sensor input for velocity calculation.

Portable Velocity Validation System

With the success of the Velocity Validation System, I was requested to create a portable and general purpose velocity validation system. I utilized similar hardware and modified the software to incorporate an intuitive graphical interface. With slim software, the system was able to achieve precise response-time within 1 millisecond and repeatable results with 2% standard deviation.

Carpet Logging System

All Roombas are rigorously tested in a controlled lab according to the IEC 62929 standards at iRobot. All carpet used for claims testing must be conditioned and logged to maintain its track record. Using Raspberry Pi, I wrote a Python script for an intuitive user interface to scan barcodes on each carpet and maintain it's data.

Mount Design

Prototype iteration of the mount design from CAD, 3D printing, and machining the PolyCarb case.

User Interface Software

The software was structured so that the graphical interface had immediate response upon use but also robust enough to execute for months with fail-safe data logging.
Powered by Create your own unique website with customizable templates.
  • Home
  • Work Experience
    • KAIST Hubo Lab
    • iRobot
    • inTouch Technologies Inc.
  • Projects
    • Embedded Computer Vision for Robot Localization
    • Vision-based Robotic Arm
    • CUAUV
    • Exoskeleton Arm
    • Robotic Segway
    • Robotic Arm Manipulation
    • Multi-core Processor
    • Piezosensor on Bipedal Robot
  • Blog