Using Machine Learning and AI to explore potential systems for costing and managing Mobility as a Service and Transport Infrastructure

使用机器学习和人工智能探索潜在的成本计算和管理移动即服务和运输基础设施的系统

• 批准号: 132994
• 金额: $ 23.73万
• 依托单位: CGA SIMULATION LIMITED
• 依托单位国家: 英国
• 项目类别: Feasibility Studies
• 财政年份: 2017
• 资助国家: 英国
• 起止时间: 2017 至 无数据
• 项目状态: 已结题
• 来源: https://gtr.ukri.org/projects?ref=132994
• 关键词: Using; Machine; Learning; AI; explore

Transport and in particular city congestion are huge issues in the 21st Century, with gridlock and pollution costing global economies billions annually. Using data sets around congestion, pollution and road safety it is possible to calculate the real cost of driving, and its impact on the environment, pedestrians, cyclists and other commuters. The project will assess the technical feasibility of a Holistic Transport Costing (HTC) Model of intelligent "pay-per-mile" micro-transactions where vehicles have to pay more to access busy or dangerous roads but are rewarded for taking more environmentally friendly routes, or carrying more passengers. The system would use machine learning (ML) and artificial intelligence(AI) to continually intelligently negotiate road usage costs between autonomous agents on behalf of the city and cars then later autonomous vehicles. The system would be capable of continual evolution to reflect priorities and behaviours in a single city, ‘evolving’ it to become less congested, more efficient, safer and cleaner. The platform would also be adaptable to multi-modal transport journeys, integrating different transport options as data around times, pricing and access became available.

交通，特别是城市拥堵是21世纪的重大问题，交通堵塞和污染每年给全球经济造成数十亿美元的损失。利用有关拥堵、污染和道路安全的数据集，可以计算出驾驶的真实的成本，以及它对环境、行人、骑自行车的人和其他通勤者的影响。该项目将评估智能“每英里付费”微交易的整体运输成本计算（HTC）模型的技术可行性，其中车辆必须支付更多费用才能进入忙碌或危险的道路，但会因采取更环保的路线或运载更多乘客而获得奖励。该系统将使用机器学习（ML）和人工智能（AI），在代表城市的自动驾驶代理与汽车之间持续智能地协商道路使用成本，然后是自动驾驶汽车。该系统将能够不断发展，以反映单个城市的优先事项和行为，使其“发展”成为不那么拥挤，更有效率，更安全和更清洁。该平台还将适应多式联运旅程，随着时间、定价和接入数据的可用，整合不同的运输选择。