DevDay

🎟️ Please RSVP here to join-in for this conversation in-person.
(Note: To ensure a spot, please register through the link)

Topic 1

From Point Sample to Gaussian Depth Field: The Math Behind Tire Tread Intelligence

A tread depth gauge returns one number: millimetres at one point, read by one person, a few times an inspection cycle. The interesting question is not how to automate that number but why we accepted it as the measurand.

This session is about that reformulation. We start from the metrology: a gauge reading is a point sample of a continuous depth field defined relative to a curved tire tread surface, so the first step was writing the measurement down properly through surface reconstruction, robust cylinder fitting for the reference geometry, and separating rib from groove as a mixture problem, not a threshold. From there the estimation problem is tractable: scale recovery in a monocular pipeline, an explicit error budget, and asymmetric loss, because under-reading a worn tyre and over-reading one are not equally expensive mistakes.

The part that changes the answer is statistical. The gauge itself is a noisy instrument, so we qualified ours the same way and treated the output as a distribution rather than a point, using conformal intervals. Once measurements are cheap and repeated, a sequence of imprecise readings tells more about a surface than a rare precise one.
Expect an honest error budget, and the wear diagnostics the gauge never gave us.

Topic 2

The AI Reshuffle: Engineering Discipline in the Age of AI-Assisted Software Delivery

Software engineering has evolved practices to help us deal with complexity, uncertainty, and the cost of getting things wrong. From pairing, TDD, refactoring and continuous integration to small batches, evolutionary architecture and DevOps, these practices have shaped how we build software that is resilient, maintainable, and adaptable.

Now AI is changing one of the fundamental economics of software development: the cost of producing code is collapsing.
But does that make our engineering practices less relevant, does it change where we need to apply them?

In this talk, I’ll revisit some of the engineering practices we’ve spent decades refining and explore how AI reshuffles their role.

🗓️ Agenda:

• 10:00 AM - 10:15 AM - Registration & Networking
• 10:15 AM - 11:15 AM - From Point Sample to Gaussian Depth Field: The Math Behind Tire Tread Intelligence
• 11:15 AM - 11:30 AM - Coffee Break & Networking
• 11:30 AM - 12:30 AM - The AI Reshuffle: Engineering Discipline in the Age of AI-Assisted Software Delivery
• 12:30 PM onwards - Lunch & Networking

Parking Notice: Vehicle parking at the venue is subject to a fee, payable by the participating individual.

Seats are limited! To ensure a great experience for everyone, please register in the link, and we’ll send out official confirmations.