The Operating Verification System
Environmental Problems. Measurable Solutions.
TOVS develops environmental monitoring methods to classify rainfall conditions, examine data gaps, and help organizations test detection performance with measurable results.
Discuss a small pilotThe challenge we address
Recognizing rainfall events when environmental measurements are incomplete or unusual
The Problem
Rainfall monitoring can involve incomplete, delayed, or unusual measurements. These conditions can make it harder to recognize events that deserve closer attention. Whether this affects a particular organization must be established from its own data.
Our Solution
Our mathematical and algorithmic workflow classifies rainfall conditions as NORMAL, WATCH, or ALERT, flag potentially significant events, and identify measurements requiring review.
Research completed: UMMDA has exercised a rainfall-classification workflow using synthetic measurements. The next milestone is independent testing on real-world measurements.
Our next step: Work with an environmental partner to test real measurements against an agreed baseline, measuring missed detections and false alerts. Discuss a small paid validation pilot.
Research example · November 2026
From a rainfall-monitoring problem to a working experimental test
The Problem We Saw
Environmental measurements can be incomplete or difficult to interpret. We investigated how to consistently distinguish ordinary rainfall readings from readings that merit closer review, without treating a measurement alone as proof of flooding.
Our Solution
We developed and ran a Python-based experimental classification workflow using three understandable outputs: NORMAL, WATCH, and ALERT. The approach provides a repeatable starting point for testing environmental event screening.
What we did: In November 2026, a local pilot processed 10 synthetic rainfall records with timestamps and expected classifications. The initial workflow reported PASS_INITIAL. This demonstrated the local test workflow; field accuracy remains to be measured.
What comes next: Evaluate real-world rainfall or stormwater records against agreed reference labels, quantify missed events and false alerts, and report whether the approach performs better than a simple baseline.
UMMDA research foundation
Seven Core Research Areas
These seven areas underpin our environmental research and focused validation studies.
1. Intellectual Property
Research concepts, methods, and associated intellectual-property development.
2. Algorithm Development
Design and evaluation of methods for interpreting environmental measurements.
3. Pre-Artificial Intelligence
Foundational analytical methods and preparation of data and logic for potential AI applications.
4. Quantum Computing
Quantum-computing research and evaluation of environmental data applications.
5. Data Integrity
Checks for consistency, provenance, and reliability of research inputs and outputs.
6. Mathematics
Mathematical modeling, measurement, and objective comparison of results.
7. Missing Data
Identifying data gaps and testing approaches to handling incomplete observations.
Focused environmental validation
A focused paid validation pilot
1. Define
Agree on one monitoring problem, a limited dataset, and success criteria before running the study.
2. Compare
Run an experimental method alongside a basic or existing approach. Record missed events, false alerts, and data limitations.
3. Decide
Deliver a concise report describing what worked, what failed, and whether further work is justified.
Research status: Our November 2026 pilot used synthetic data. Real-world performance, including missed events and false alerts, will be measured in subsequent validation. Experimental outputs are not a substitute for established safety-critical warning systems.