INRIX has expanded its predictive parking and curb intelligence platform, covering 48 million parking spaces across 22,000 cities in 145 countries and drawing on data from 300 million connected vehicles. The platform predicts parking availability at estimated time of arrival, with demand forecasts extending up to seven days ahead. The platform is validated through monthly bench testing and quarterly real-world testing across cities including Berlin, London, Los Angeles, and Boston. Between January and April 2026, testing across 46 city instances in 12 countries delivered approximately 99% coverage accuracy, with attribution accuracy of at least 90% on driver-facing data including pricing and parking rules. For automotive OEMs and mobility platforms, the data supports parking-aware routing, autonomous arrival planning, and reduced congestion from unnecessary cruising near destinations. The platform draws on more than 44 billion data points generated daily, with predictive models trained on trillions of connected vehicle events. In a statement, Ahmed Darrat, Chief Product Officer at INRIX, said: “Parking intelligence is no longer about static maps. It’s about helping drivers, vehicles, and cities make confident decisions before arrival based on real-world conditions, regulations, and probability of availability.” Darrat added: “The future of mobility depends on knowing, not guessing, what will be available when you arrive. INRIX is defining that standard with predictive, validated, and globally deployed parking intelligence.” Why this matters: • The autonomous vehicle dependency is the least-discussed constraint in AV deployment planning. A self-driving vehicle operating without a human to circle the block needs high-confidence parking availability data before it departs, not on arrival. INRIX’s predictive models, trained on connected vehicle data and forecasting up to seven days ahead, address a precondition for AV commercial viability that the hardware and software announcements from GTC Taipei this week largely took for granted. The infrastructure layer beneath autonomy is broader