Building on the v14.2.1 refinements from the December 13 Holiday Update, Tesla has begun deploying Full Self-Driving (Supervised) version 14.2.2. This update enhances obstacle detection, emergency vehicle responses, and introduces customizable arrival options. Early users report smoother, more confident performance across diverse conditions.
Tesla initiated the rollout of Full Self-Driving (Supervised) v14.2.2 on December 23, 2025—following the minor v14.2.1.25 tweaks in the prior Holiday Update—prioritizing smoother real-world driving and advanced obstacle awareness. Key upgrades include a higher-resolution vision encoder neural network for better detection of emergency vehicles (police, fire trucks, ambulances), road obstacles, and human gestures. New Arrival Options let users choose drop-off spots like Parking Lot, Street, Driveway, Parking Garage, or Curbside, dynamically adjusting the navigation pin.
Enhancements encompass yielding or pulling over for emergency vehicles, integrating navigation into the vision neural network for real-time detours, and improved handling of gates, road debris (tires, branches, boxes), unprotected turns, lane changes, cut-ins, and school buses. Reliability boosts feature fault recovery, interior windshield residue alerts, and auto narrow-field camera washing for 2026 Model Y.
Early feedback is enthusiastic: @BLKMDL3 noted zero lane-change hesitation over 10 hours, refined parking, and strong dark canyon performance. Dan Burkland praised 'buttery smooth' nighttime drives akin to Robotaxi. Devin Olsen aced heavy traffic and rain. Xpeng CEO He Xiaopeng hailed it as near-Level 4, far surpassing prior year, crediting Tesla's unified software.
These steps advance FSD toward unsupervised use, with future emphasis on smoothness and parking.
