Mobility as a Service (MaaS) systems are the digital platforms which aggregates the various transportation options available between any points and helps the users to plan and manage their trips seamlessly and efficiently. MaaS is getting wide acceptance and is now being used by many countries and organizations across globe. A report by McKinsey & Company projected that by 2030 the global MaaS market could be somewhere between $500 billion to $1.5 trillion. Like in another areas, artificial intelligence (AI) plays a significant role in the development and evolution of MaaS systems too.

AI Adoption in MaaS

AI can play a significant role in MaaS systems.

• Data analysis and prediction: MaaS system relies on huge amount of data from heterogenous sources such as transportation schedules from different providers including land, water and air transportation systems, real time traffic and weather conditions, user preferences etc. AI algorithms can be applied for analyzing this data and presenting an optimal route based on the user preferences and the physical conditions.
• Demand forecasting: AI based models can be used to analyze the historical data patterns and forecast the future demands for transportation needs and services. This helps the transportation agencies and the MaaS operators to allocate the resources efficiently.
• Personalized recommendations: AI algorithms to learn from the user preferences such as preferred modes of transportation, preferred routes, departure / arrival times etc would help to make recommendations which are must suitable for the user based on his choice.
• Promotions and campaigns: By analyzing the demand, availability and other factors, the AI models can help the transport agencies to propose incentive / reward programs during off peak hours to have an optimized usage of resources and routes.

Future of AI in MaaS

As MaaS systems get widespread acceptance, AI should be able to play a wider role in terms of:

• Predictive maintenance: By analyzing data from sensors, vehicle diagnostics, and historical patterns, AI algorithms can proactively identify potential maintenance needs, reducing downtime and ensuring the reliability of transportation services.
• Sustainability and environmental impact: AI can contribute to the development of more sustainable systems by optimizing the routes, reducing idle times, and promoting the use of eco-friendly transportation modes and can help accelerating the shift towards greener options.
• Integration of emerging transportation modes: As new transportation modes like autonomous vehicles, drones, and hyperloop systems become more prevalent, AI will play a crucial role in integrating these modes into MaaS systems. AI algorithms will help optimize the coordination, scheduling, and integration of these emerging modes to provide seamless multi-modal experiences.
• Natural Language Processing (NLP): NLP models can be employed to process and understand user queries and commands. This enables MaaS platforms to provide conversational interfaces, voice-based interactions, and enhanced user experiences.
• Advanced analytics and decision-making: AI will continue to improve analytics capabilities in MaaS systems, enabling more sophisticated analysis of transportation data. This can help MaaS operators and urban planners make data-driven decisions for optimizing transportation networks, infrastructure planning, and policy-making.

Challenges of AI in MaaS

Though MaaS systems can benefit quite a lot with AI adoption, there are its own challenges too and a few are:

• Privacy and security: Since the AI modules in MaaS systems involves collecting and processing huge amount of personal data such as travel information, modes of travel etc, the AI system must comply to the PII guidelines and ensure that there are no data breaches or misuses happens.
• Data Quality: As MaaS systems are still in evolution phase, the quality and quantity of the data is a major concern for developing efficient and accurate AI models.
• Scalability and integration: Developing AI models for MaaS could be relatively complex as it involves integration across different systems like transportation providers, route managing, payment systems, user preferences etc.

A comprehensive design approach for AI models considering the data privacy guidelines, geographic and regions specific laws for data protection, complexities in the systems involved and integrations etc are inevitable for a successful AI adoption in MaaS systems.

An insight of MaaS providers utilizing AI

There are many MaaS providers who started utilizing the AI capabilities for enhancing their services and offerings. Here are a few examples:

• Citymapper: City mapper applies AI technologies to analyze transit data, predict delays, suggest optimal routes and personalized recommendations based on historic data patterns.
• Moovit: Moovit uses AI to aggregate and analyze vast amounts of real-time data from various transit agencies, to provide accurate information on bus and train schedules, delays, and route changes.
• Lyft: Lyft utilizes AI algorithms for ride matching, dynamic pricing, and ETA predictions. AI is also employed for fraud detection and prevention to ensure the safety and reliability of their services.
• SkedGo: SkedGo partnered with AI Base Technologies from Japan to provide an AI based MaaS solution with deep data analytics and machine learning capabilities.

Well modelled AI systems can play a major role in Route optimization, Fare optimization, Personalized trip recommendations, Demand Prediction, Real time data analytics, Predictive maintenance and several other areas.