Sandra Mau Master of Business Administration Candidate at Queensland University of Technology Master of Science in Robotics at Carnegie Mellon University 2007 Bachelor of Applied Science in Engineering Science (Aerospace Engineering) at University of Toronto 2005 My Technical CV [Australia pdf] [Australia doc] [North America pdf] (Available upon request) My Robot Portfolio [pdf] My CMU Robotics Profile My Business CV [Australia pdf] [Australia doc] (Available upon request) Email: smau@andrew.cmu.edu

Research Interests

• Social application of technologies
• Robot localization, navigation and planning
• Machine learning and inference
• Human robot interaction
• Intelligent user interfaces

Labs I have worked with

• Tele-Supervised Autonomous Robotics (CMU)
  Research in the T-SAR lab focuses on providing end-to-end tools for tele-supervision of autonomous robots.
  Heads: Gregg Podnar and John Dolan
  
  
  • Wide Area Prospecting Using Supervised Autonomous Robots Project
    Together with NASA's Jet Propulsion Laboratory and the NASA Ames Research Center we will develop a software architecture and system for autonomous cooperative perception, control, coordination, navigation, and information-sharing / management of multiple robots: the Robot Supervision Architecture (RSA).
    Head: John Dolan
  
  
• Space Robotics (UTIAS)
  The focus of this research group is the space application of robotics.
  Head: Gabriele D’Eleuterio
  
  
• Ornithopter (UTIAS)
  The purpose of this project is to achieve humanity's oldest aeronautical dream: flight with mechanical flapping wings.
  Head: James DeLaurier

Publications

• S. Mau and J. Dolan, "Scheduling for Humans in Multirobot Supervisory Control," Proceedings of the International Conference on Intelligent Robots and Systems (IROS), October, 2007. [pdf] [Abstract] [Bibtex]
  
  
• S. Mau and J. Dolan, "Scheduling to Minimize Downtime in Human-Multirobot Supervisory Control," Workshop on Planning and Scheduling for Space, October, 2006. [pdf] [Abstract] [Bibtex]
  
  
• A. Elfes, J. Dolan, G. Podnar, S. Mau, and M. Bergerman, "Safe and Efficient Robotic Space Exploration with Tele-Supervised Autonomous Robots," Proceedings of the AAAI Spring Symposium, March, 2006, pp. 104 - 113. [pdf] [Abstract] [Bibtex]
  
  
• E. Halberstam, L. Navarro, R. Conescu, S. Mau, G. Podnar, A.D. Guisewite, H. Brown, A. Elfes, J. Dolan, and M. Bergerman, "A Robot Supervision Architecture for Safe and Efficient Space Exploration and Operation," Tenth Biennial International Conference on Engineering, Construction, and Operations in Challenging Environments: Earth & Space 2006 Conference, ASCE, March, 2006. [pdf] [Abstract] [Bibtex]

Theses

• S. Mau, "Scheduling for Humans and Team Size Issues in Multirobot Supervisory Control," Master's Thesis, Tech. Report CMU-RI-TR-07-20, Robotics Institute, Carnegie Mellon University, June, 2007. [Thesis pdf] [Abstract] [Bibtex] [Presentation pdf]
  Looking into various methods of efficiently utilizing human time in a multirobot supervisory control system. In particular scheduling tasks for humans which allow robots to regain autonomous control sooner and looking at the factors which determine maximum team size a human can handle before saturation.
  Thesis advisor: John Dolan
  
  
• "Experiments on the Estimation of the Fundamental Matrix Applied to a Two-View Geometry in Visual Localization," Bachelor's Thesis, April, 2005. [Thesis pdf]
  Experimental research into two-view vision geometry as a method for object detection and localization. Specifically, it takes a look at various methods of finding the fundamental matrix, which holds the information about how the two views are related. Using this information a 3D scene can be reconstructed.
  Thesis advisor: G.M.T. D'Eleuterio

Posters

• S. Mau and J. Dolan, "Scheduling to Minimize Downtime in Human-Multirobot Supervisory Control," Workshop on Planning and Scheduling for Space, October, 2006. [pdf]
  
  
• N. Armstrong and S. Mau, "Robust Assistive Navigation with the Intelligent Mobility Platform (IMP)," AI class project, December, 2006. [pdf]
  
  
• S. Mau, M. Makatchev and N. Melchoir, "BlindAid: An Electronic Travel Aid for the Blind," Quality of Life Technology Summit, November, 2007. [pdf]

Current and Past Projects

• BlindAid: Navigational Assistance for the Visually Impaired [site], May, 2007. [Final Report pdf]
  The goal of BlindAid is to develop navigational assistance technology for the blind or visually impaired. We seek to create a portable Electronic Travel Aid (ETA) for visually impaired users, along with the accompanying radio frequency identification (RFID) localization infrastructure used to equip buildings.
  Project advisor: Aaron Steinfeld
  Course advisors: Bernardine Dias, Manuela Veloso, (V-unit) Illah Nourbakhsh (Principles of Human-Robot Interaction)
  
  
• "Navigational Policy Generation for the Intelligent Mobility Platform Using Wireless and Stereo Vision for Localization," Graduate AI and Statistical Techniques for Robotics class project, December, 2006. [pdf]
  The purpose of this project was to develop navigational policies based on a fusion of sensor modalities: stereo vision and wireless signal strength. These sensors were chosen in part due to complementing strengths of local and global localization. A particle filter was used to keep track of belief points and two POMDP variants were investigated for planning. One was the Augmented MDP and the other was an E-PCA Compressed Belief MDP.
  Course advisors: Drew Bagnell (Statistical Approaches to Robotics), Geoff Gordon (Graduate AI)
  
  
• Particle Filter, Statistical Techiques for Robotics homework report, Occober, 2006. [pdf] [Video]
  The problem we are trying to address is localization in a mapped location based on laser range data using a particle filter. Specifically, we use Monte Carlo Localization, a localization algorithm based on particle filters.
  Course advisors: Drew Bagnell (Statistical Approaches to Robotics)
  
  
• Online Classification, Statistical Techiques for Robotics homework report, November, 2006. [pdf]
  Report on the use of various data classification algorithms (binary and multiclass) on several datasets. Algorithms used include Gaussian Processes, SVM, NORMA and Winnow
  Course advisors: Drew Bagnell (Statistical Approaches to Robotics)
  
  
• "Effective Designs of Non-Keyboard Computer Input Devices," Mechanics of Manipulation class project, April, 2006. [pdf]
  Literature review on limits of the human form and how exceeding it can cause musculoskeletal disorders.
  Course advisor: Matt Mason (Mechanics of Manipulation)
  
  
• "Kalman Filter and Data Fusion," Robotics Math class project, December, 2005. [pdf]
  Report of experiments on the integration of IMU and GPS information through various Kalman filter techniques to improve heading.
  Course advisor: Mike Erdmann (Math Fundamentals for Robotics)
  
  
• "Hubble Space Telescope Robotic Servicing Mission Design Report," Space Systems Design project report, December, 2004. [pdf]
  The Hubble Space Telescope has been in service since 1990. Currently, the Hubble is approaching a point where significant maintenance is required. Its ability to perform scientific research has been hindered by aging power sources and it may loose the ability to stay in orbit by 2009. NASA’s proposed servicing mission involves the use of a launch vehicle called the Hubble Robotic Vehicle (HRV). It will contain a De-orbit Module, Ejection Module, and two robotic manipulator systems: Grapple Arm (GA) and Dexterous Robot (DR). The purpose of this launch vehicle and its robotic systems is to avoid Extra Vehicular Activity (EVA) performed by astronauts and the use of a space shuttle. The focus of this report will be on the design of the GA.
  Course advisors: Paul Fulford, Tim Reedman, Ross Gillett, Tim Fielding and Perry Newhook ( MacDonald, Dettwiler and Associates Ltd. - fomerly known as MDRobotics); Chris Damaren (University of Toronto)
  
  
• "Kwik-E-Kart: The Robot Waiter," Engineering Design Project (draft), March, 2003. [pdf]
  This design project involved undertaking the construction of a machine to accomplish certain tasks. On a black and white grid half the size of a tennis court were located a maximum of seven randomly placed 6-inch square metal plates. The objective of the machine was to follow an inputted path of nodes, while detecting the plates situated on the lines and depositing a pop can vertically onto the plate. Moreover, the machine was not allowed to come into contact with any plates, and had to have the capability of being loaded easily with a maximum of seven cans for dispensing purposes. In essence, the machine was to act as a waiter. Note: not all sections are included in the pdf (in particular the appendices).
  Course advisor: Dr. Reza Emami (Engineering Design)
  
  "Ornithopter Wing Optimization," Ornithopter summer internship report, August, 2003. [pdf]
  A new ornithopter wing was designed using analytical software to theoretically produce enough lift and thrust to propel an engine-powered piloted aircraft into steady flight.
  Research advisor: James DeLaurier (on Wikipedia!)
  
  

Other Interests

• Entrepreneurship
  • Received Entrepreneurship, Innovation and Small Business Certificate from University of Toronto, May 2005.
  • Participated in University of Toronto Engineering Kompetition for Entrepreneurial Design (placing second in both 2003 & 2004) and Ontario Engineering Competition
  
  
• Ultimate Frisbee - with Money Mellons, CMU women's team
  [Money Mellons Team Pic! (I'm top left)]
  
  
• Movies - there's always time for a good movie!
  Top 5:
  1. The Shawshank Redemption
  2. Dead Poet's Society
  3. Forest Gump
  4. Lord of the Rings (Return of the King especially)
  5. Shrek
  
  
• Dancing - Various Latin and Ballroom. Gumboots! [pic]
  
  
• Volunteering
  
  • Creative Technology Nights for Girls
    Creative Technology Nights for Girls is a program focused on exposing middle and high school girls to creative technologies. Using computer animation, web design, programming, robotics, and interactive medias, we hope to engage a future generation of women in technology.
  • Animal Rescue League
    Helping to socialize cats and dogs and showing them to potential adopters.
  • Canadian Cancer Society
    Outreach for cancer awareness and prevention. Fundraising through events and activities. Maintaining records of people using the assistance programs provided by Cancer Society.
  • Math teacher for children through Tropicana Community Services
    A not-for-profit organization providing culturally appropriate social services in east Toronto, focusing on the needs of youth, and the Caribbean and Black communities.
  
  
• University of Toronto AI and Robotics Club (formerly known as UofT RoboCup)
  Competitve robotics for competition. Previous ones include Eastern Canadian Robot Games and Trinity Fire Fighting Robots.