What if the future of transport is already here? The Bayesian model predicting electric robotaxi adoption

Electric Level 4 automated vehicles (CAVs) are already being tested as robotaxis in several cities, driving autonomously within geofenced zones. Policymakers and companies need to know when people will start using them. Researchers built a Bayesian predictive model to forecast CAV adoption from 2030 to 2035. The model takes effectiveness estimates of CAV applications and runs Monte Carlo simulations to capture uncertainties in technology and infrastructure readiness. Scenarios show that CAVs are likely to be adopted if technological progress continues and users can experience their capabilities in real‑world services. Reliable demonstration results significantly raise adoption probability. The findings guide governments and industry groups in preparing regulations and infrastructure for a greener, automated transport future.

The Bayesian framework integrates effectiveness estimates, prior distributions, and Monte Carlo sampling to forecast electric Level 4 CAV adoption. Inputs comprise metrics such as cost per mile, user satisfaction, and charging infrastructure maturity, each represented as a probability distribution to capture inherent uncertainty. Monte Carlo simulations propagate these uncertainties, producing a posterior distribution of projected adoption rates for 2030‑2035. Multiple scenarios are evaluated, including rapid technological advancement, expanded charging networks, and high‑reliability demonstration programs. The analysis quantifies the value of information, revealing that each credible demonstration reduces uncertainty and pushes adoption probability above a critical threshold. This method enables policymakers and industry stakeholders to pinpoint the most impactful investments for accelerating CAV integration into public and private transport fleets.