<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=984453251645033&amp;ev=PageView&amp;noscript=1">

Using Machine Learning to Make Shift Reports Smart

Learn how we use machine learning to improve the quality and indexing of shift reports for organizations and why machine learning is needed for the future of aquatics management.

At DigiQuatics head guards, managers-on-duty, pool managers, and supervisors use DigiQuatics Shift Reports or "end of shift reports" to communicate how the day-to-day operations at the pool are going. This could include chemical issues, staffing issues, patron complaints, issues in the pump room, or even an incident report with documentation of what a lifeguard did to help a child that fell and scraped his knee.

Before DigiQuatics, this was all done on pen and paper. Now, having everything handled digitally and available for search has been a huge help for organizations, but we had to ask ourselves, "what is the next step in this evolution?" Well, a solution that leverages machine learning (ML) to constantly improve how reports get tagged, indexed, and dive deeper into the content of uploaded incident reports and accompanying documents (often containing PDFs or images). So, let's talk about why machine learning is needed for an aquatics app, and how do we use it?

Why Machine Learning?

The way you can determine the success of end-of-shift communications is the ability to analyze, summarize, and index this content. The more indexed a shift report is the easier it is to find and sort through all of the hundreds or even thousands of reports to find exactly what you need in the click of a button. DigiQuatics leverages multiple machine learning (ML) services to provide this experience and service.

Before we dive into the services and problems we solve with ML, let's define a few terms.

Sentiment analysis refers to the use of natural language processing, text analysis, computational linguistics, and biometrics to systematically identify, extract, quantify, and study affective states and subjective information.

DataModel-New-Page.png

Optical Character Recognition (OCR) refers to both the technology and process of reading and converting typed, printed or handwritten characters into machine-encoded text or something that the computer can manipulate. This feature is coming soon to the Shift Reports module.

Screen Shot 2017-11-22 at 4.43.44 PM.png

Key Phrase Extraction extracts key phrases from text to quickly identify the main points. For example, for the input text ‘There is a problem with the pool pump' this specialized ML service would return "pool pump" as a key phrase in a shift report.

Untitled Design.png

Shift Report Show (1).png

Example Shift Report from DigiQuatics

How Does It Work?

Our team utilizes multiple ML services to provide enhancements to the Shift Reports module in DigiQuatics. The key phrase extraction and OCR services are constantly learning and evolving as more data gets run through them. It's a constantly evolving system. Generally, sentiment analysis is performed on a purely algorithmic basis, using weighted scores for various words in the text to generalize the overall sentiment in a shift report.

Closing Thoughts

We hope this blog post was insightful about where we are at with DigiQuatics and where we are going in the future. Right now, this is just the beginning. Our team plans to leverage multiple algorithmic and ML techniques and models to take aquatics management to the next level. Some of the applications include auto-scheduling, predicting busy days for pools using weather forecasts, patron counts, and historical data, and also predicting what the chemical levels in a body of water will be given historical chemical records, patron count data, and more!

Comment below with an idea on how you think machine learning could be used in your everyday pool operations!






Subscribe to DigiQuatics Blog