Session: 16-01-01: Government Agency Student Poster Competition

Paper Number: 150809

150809 - Automating Chain of Custody Documentation for Filter-Based Air Pollution Sampling 

Air pollution is a leading risk factor for death worldwide, and efficient, economical, and accurate air pollution measurement can be an effective means of creating impactful change but is also challenging. In recent years, increased access to air pollution sampling systems, filter-based sampling in particular, has empowered communities to measure, analyze, and take action to protect their air quality. However, filter-based air quality data is only as usable as our confidence in how and where those samples were collected. Ensuring a proper Chain of Custody (CoC) for air pollution samples is crucial for generating reliable, accessible data and is often legally required. Unfortunately, filter-based air quality measurements are often time-consuming and labor-intensive. The limited availability of personnel and the tedious nature of documenting filter CoC information can result in human errors, rendering many samples unusable and limiting the scope of available data. Automating the CoC process would allow for easy and efficient documentation, improving data quality while simultaneously allowing personnel time to be dedicated to more important activities. An effective automated chain of custody system must be able to recognize and record filter identifiers, log this information along with the time and location of the filter, and physically organize the filters for future transportation or storage. The system must also be able to handle filters gently to prevent damage, identify and separate any incorrectly scanned filters, and provide the user with documentation of the entire process. The solution I have developed is an electro-mechanical system called the Chain of Custody Automation Bot (CoCab). CoCab automates sorting and recording filters to allow for a digital chain of custody with minimal user input. Each filter is mechanically moved into a scanning station equipped with two oppositely mounted cameras to allow filter identification regardless of loading orientation. Computer vision is used to decode filter identifiers; successful filters are moved to temporary storage for user collection, while unsuccessful filters, those for which either an ID was not found, or the ID did not match predetermined requirements, are moved to a secondary location for user review. CoCab runs a custom Python-based program that and is operated by a Raspberry Pi 4 B+ microcomputer controlling stepper motors and cameras and reporting progress to a graphical user interface. Creating a Chain of Custody (CoC) for a 10 filters currently takes approximately 5 minutes. Upon full integration of CoCab, the time required for personnel to document the CoC for the same number of filters will be reduced to 1 minute. This reduction in personnel time required for chain of custody documentation will help further streamline the process of collecting air pollution data and render data from a higher portion of collected samples usable. Ultimately, this will help increase the amount of meaningful, high-confidence data and data-driven change.

Presenting Author: Apurva Iyengar Colorado State University

Presenting Author Biography: Apurva Iyengar is 2024 National Science Foundation REU participant through the Airborne Connections REU at Colorado State University, advised by Dr. Christian L'Orange. She is also an undergraduate student at Tufts University pursuing a degree in Mechanical Engineering and minoring in Geology and Math. Apurva is deeply passionate about climate and air quality issues and has been a strong advocate for environmental causes since high school.

Authors:

Apurva Iyengar Colorado State University
Christian L'orange Colorado State University