Supervised assist
Driver-responsible systems (e.g. FSD Supervised) generate miles and training signal — not the same as driverless commercial service.
Autonomy · Evidence · Humility
Framing how we think about autonomous driving safety for a Hamilton fleet concept: data loops, supervised vs unsupervised distinctions, and the refusal to overclaim Canadian readiness.
Qualification notice
Safety statistics published by Tesla or third parties for other markets or for FSD (Supervised) do not automatically transfer to unsupervised commercial robotaxi service in Ontario. This page does not assert that Cybercab operation is currently authorized or statistically proven for Hamilton roads.
Driver-responsible systems (e.g. FSD Supervised) generate miles and training signal — not the same as driverless commercial service.
Limited, permitted operations under evolving Canadian / Ontario frameworks — may require attendants, ODD limits, or reporting.
Commercial robotaxi without a safety driver is an aspiration gated by Tesla capability, regulators, and insurers — timeline unknown.
Software-defined autonomy improves through fleet data, simulation, and staged release. Local operators still owe riders operational safety: incident response, first-responder access plans, maintenance quality, and willingness to park vehicles when conditions exceed the operational design domain.
Camera-heavy neural approaches emphasize scalable sensing and software iteration. Weather, glare, and edge cases remain active research and validation problems — especially in Canadian winter conditions.
Other AV programs combine lidar, radar, and cameras. We do not claim superiority of any architecture here — only that Cybercab planning inherits Tesla's chosen stack and its tradeoffs.