MTA to Use New Simulation Technology to Train Bus Drivers
google openai
| Source: Mastodon | Original article
The Metropolitan Transportation Authority has commissioned a $1.4 million, AI‑enhanced bus‑driving simulator that will soon replace the bulk of on‑road training for new operators. Installed at the Zerega training centre in the Bronx, the system projects realistic traffic, weather and passenger‑load scenarios on a full‑scale virtual replica of Manhattan’s streets. Trainees can practice lane changes, stop‑sign compliance and emergency braking without endangering commuters or wear‑and‑tear on the fleet.
The move arrives amid a chronic driver shortage and a spate of safety incidents that have pressured the MTA to modernise its onboarding pipeline. By front‑loading skill acquisition in a controlled digital environment, the agency expects to cut the time to certification by up to 30 percent and reduce early‑career crash rates, according to officials. The simulator’s AI core, built on models from OpenAI and Google AI, generates dynamic traffic patterns that adapt to a driver’s decisions, offering a level of variability that static video‑based courses cannot match.
Industry observers note that the technology mirrors what airlines have used for decades and what autonomous‑vehicle firms are testing today. If the pilot proves successful, the MTA plans to roll the system out to its other depots and to integrate real‑time service data, allowing trainees to rehearse disruptions such as construction detours or severe weather events. The agency also hinted at a future “digital twin” of the entire bus network, where AI could simulate fleet performance under different policy scenarios.
Watch for the first batch of graduates emerging from the simulator later this year, and for the MTA’s post‑implementation report, slated for early 2027, which will reveal cost savings, safety impacts and whether other transit authorities will adopt similar AI‑driven training platforms.
Sources
Back to AIPULSEN