Discovery Park Undergraduate Research Internship Program

"Autonomous Indoor Localization for Fire Safety and Resource Location via Field Mapping Techniques"

About the Project

Project Time & Type:
Fall 2014 - DURI
Research area(s):
Mechanical engineering
Project Description:
We have shown cell phone geolocation to within 20cm indoors using a magnetic field map, and the magnetometer and inertial sensors available on all smart phones. Figures 1 and 2 sum up our results for a corridor in the Purdue ME building. Our method [2,3] is novel and uses the disturbances in the indoor magnetic field vectors—caused by iron in construction, furniture and appliances—as beacons for pin pointing position. Our method is applicable to areas where the magnetic map does not change faster than it can be mapped, i.e., any indoor space where spatial concentrations of ferrous materials remain static such as libraries, offices, or hotels. Finally, our algorithms have clear mathematical properties in that geolocation errors depend only on the resolution of the sensors and the size of magnetic disturbances in the environment.
Expected Student Contributions:
Students will become familiar with aspects of research volunteer scheduling and study assistance, and support for a 3-5 day data collection of cell phone mapping in the Potter Library where a crowd-sourced map from collective efforts will be generated. After data collection phase, student(s) will gain a full and in-depth data curation lifecycle experience that includes gaining data science skills that involve implementing research code on data sets, analyze statistics for map convergence, and check on the software running smoothly. Cui Yan, and Dewayne Branch will develop a work flow of cell use training, data collection, data storage and data curation for this Stage 2 effort of indoor mapping. Student(s) will work 6-8 hours on this project a week and train in the prior weeks before cell phone data collection efforts and work towards a successful data curated project for Stage 2 research implementation. Study participant populations will consist of both students and emergency personnel. All data collected will be in an anonymized form for later review, publication, analysis and map building.
Related Website(s):
Desired Qualifications:
Mechanical Engineering Students
Estimated Weekly Hours:
Department awards independent research credits for this project?
Yes, 3 credit hours

Professor in Charge

Branch, Benjamin

Student Supervisor

Benjamin Branch
CLIR/DLF Data Curation Fellow

Cooperating Faculty

Kartik Ariyur
Mechanical Engineering
Cui Yan
Mechanical Engineering