Dev By Brice

06/17/2026

One of the most common assumptions in infrastructure management is that older assets are inherently at greater risk of failure.

While age certainly matters, my experience has been that the reality is often much more complicated.

Two pipes installed in the same year can perform very differently depending on factors such as material, installation practices, operating conditions, soil environment, maintenance history, and system demands. Age tells part of the story, but rarely the entire story.

As utilities increasingly adopt data-driven asset management strategies, understanding the difference between age and risk becomes even more important.

I recently wrote an article exploring why asset age alone is often a poor predictor of infrastructure failure and why a broader risk-based approach can lead to better decision-making.

Read the full article here:

https://www.devbybrice.com/technical-journal/ml/why-asset-age-is-a-poor-predictor-of-infrastructure-failure

06/07/2026

I have made quite a bit of progress on my Hydraulic Analysis Suite project. I have the Pipe class written and tested it out.

I tested the results by creating an instance in a jupyter notebook and printing out a summary method which contains all of the attributes and properties I have created. I tested the computer generated results against hand calculated results. This isn't any difficult calculations, but you still need to make sure of no bugs.

All calcs checked out. What does need revisions is the hydraulic properties. I had AI generate the hydraulic properties, then called on those properties in a modular file (using a dictionary for you computer nerds -- Later, I will store in a db but for now it is a simple dictionary). Those values are not lining up with industry standards. Probably fine for a college course, but I am not building for that level of accuracy. So, I will need to likely add a lot more entries and refactor, but I know my property calcs are spot on!!

06/06/2026

Currently, I'm working on a project I'm calling Hydraulic Analysis Suite.

The goal is to build a municipal engineering software package in Python, starting with pressure pipe headloss calculations using both the Hazen-Williams and Darcy-Weisbach methods.

If those terms mean something to you, then you've probably spent some time around water, wastewater, or utility infrastructure projects.

This project isn't about machine learning or data science. It's about strengthening my software engineering skills while building something directly connected to the work I've done throughout my engineering career.

I'm writing the application from scratch in Python, focusing on clean design, engineering calculations, testing, and maintainability.

It's still early in development, but it's been a fun challenge combining civil engineering and software development into a single project.

05/09/2026

🚤 Recently helped enhance Stone Transport's website; improved the booking flow and polished the UI. Check it out if you’re looking to transport your boat or just curious about smooth online experiences!

💡 Bonus: Reach out directly for transport inquiries.

Nationwide professional boat transport services. Reliable, insured, and experienced in handling all vessel types. Get a free quote today!

05/09/2026

🚀 Long time no post! I’ve been heads-down building BudgetSync, a backend first budgeting app designed for real-world financial tracking. Next step: deploying it on Render with PostgreSQL and Redis, excited to see it fully live.

Love hearing from fellow devs or finance nerds: what features would you want in a smart budgeting tool? 💡

Check out more at my website: https://devbybrice.com

Applied machine learning and backend engineering focused on real-world data, production systems, and scalable software.

01/19/2026

I’ve started an applied machine learning project focused on water main break risk, using public asset data from the City of Toronto.

What interested me wasn’t prediction accuracy, but the role ML can realistically play in infrastructure decision-making. In many U.S. cities, asset-level water main data suitable for modeling simply isn’t publicly released, which makes certain types of analysis impossible regardless of technique.

Toronto stood out because it publishes asset-level data in a form that’s usable for modeling, not just mapping.

This project is in the early-stages and exploratory phase. I’m documenting the approach, constraints, and tradeoffs as it develops, with an emphasis on decision support rather than abstract metrics.

The call came in this morning: a pipe break at a treatment plant. Crews were already managing the immediate impacts, but the real story…