Inductive loops have served networks well, but maintainers experience the trade-offs: they are vulnerable to pavement work and can be costly and disruptive to renew. Meanwhile, AI-enabled roadside sensing has matured rapidly, creating options that may be installed and iterated with less disruption. ASM is trialing alternatives to traditional detection, including: AI video analytics sensors that derive counts and classifications from video, in‑ground magnetometers for loop replacement, and radar detection to improve resilience where loops are high-risk or hard to maintain. These trials are being run from the maintainer's perspective: not 'can the technology work in a demonstration', but 'can it be installed safely, calibrated consistently, maintained predictably, and trusted by operators.

This subject is topical and timely because agencies are being asked to deliver smarter transport outcomes while minimising disruption and improving safety and reliability. For ASM, detection is not a single device decision; it is a whole-of-life system decision involving accuracy, lifecycle cost, safety and maintainability. This presentation is relevant to T‑Tech 2026 because it translates emerging technologies into the practical questions the sector needs to answer: where should we replace loops versus retain them, what evidence should be required to 'promote' pilots into BAU, and how can we standardise evaluation so results are comparable across regions and suppliers. We will share progress to date and invite discussion on what 'good enough to operate' means for New Zealand networks.