The cleanup of military munitions requires estimating site contamination, documenting decisions and actions, and monitoring the data used to make the decisions. Tremendous success and advancements have been realized by Department of Defense researchers and consultants using advanced geophysical classification (AGC) methods during the past decade. The implementation of AGC is complex, detailed-orientated, and requires trained data analysts. Broadband electromagnetic induction (EMI) data are acquired by complex and intricate sensors. Once acquired, the raw AGC EMI data must be processed using sophisticated data analysis software to make classification decisions.
The focus of this project, UX-Classify, was a cloud-based EMI data classification software suite designed to facilitate transparent program flow, efficient processing, and excellent classification results. The primary technical objective of this project was to demonstrate and evaluate the efficacy of using UX-Classify in support of military munitions remediation program (MMRP) projects that utilize AGC.
Two additional objectives were to (i) develop designs and requirements for adding capabilities for processing EMI data collected while the sensor is moving (viz., survey mode), and (ii) enhance the simulator codes to include common failure modes that can occur during data collection.
UX-Classify is a cloud-based classification and project management application for the MMRP community. UX-Classify combines the benefits of cloud computing with proven EMI data inversion methods. The software suite utilizes data inversion to extract and classify features intrinsic to buried metallic objects.
UX-Classify is composed of four components: a web application, an application program interface service, data storage, and a processing engine. The advantages of this technology are the result of a unique project management approach, strict adherence to the AGC quality assurance project plan, and the revolutionary achievements of cloud technology. Specific advantages include:
Four experienced persons were asked to participate in a demonstration and provide feedback: senior data processors from a commercial firm (data analyst and QC), a geophysicist, and an active state regulator that serves on ESTCP’s AGC technical advisory committee. Although all four were experienced with MMRP technologies, processes, and goals, none of them had previous experience with, or detailed knowledge of, the UX-Classify software. The contractor presented each participate with a two-hour introduction and provided access to a pre-demonstration data product so that they could gain familiarity with the software if they desired. All four project participants provided positive remarks and/or constructive criticism. Given the radical change in user experience that UX-Classify offers, the user responses were very favorable. The software was considered easy to use, it required minimal training, it loaded without issue on government computers, and it provided unprecedented access for the quality assurance and regulator into the data processing and decision-making phases of the AGC process. In fact, the ease of use was rated 10 out of 10 by two of the participants.
The government created a six-acre site with 2,000 plus sources (not including QC data). The simulator software collected 180 cued collections per day, which resulted in data being collected over a 13-day span from 1-23-2022 to 2-4-2022.
All quantitative performance objectives were satisfied:
Because the data set resided in cloud infrastructure instead of local drives or networks, data transfer between parties was realized by changing data access permissions within UX-Classify. As a result, data transfer issues that plague networked personal computer-based schemes were eliminated.
In compliance with this task, a dynamic data processing flow was designed that seamlessly integrated with, and leveraged, the UX-Classify (cued) application. Due to the complexity of the workflow, the design included many options and variables. As a result, and in partnership with ESTCP, the contractor has been contributing internal resources to incrementally implement the design. As a result of this collaboration, the basic framework has been implemented. There are no known technical implementation issues that will prevent this technology from working.