How AI can help overcome challenges in the transport sector

At Oaklin, we recognise the challenges an ever-evolving technology landscape brings to the transport sector. Integrating modern technologies whilst maintaining consistent travel experiences can be extremely challenging. It calls for thorough decision-making and innovative project management. 

Many of the organisations we spoke to at the Transport AI 2024 event in January questioned the effectiveness of current procurement practices for AI solutions. The overwhelming consensus was that AI suppliers should carefully develop solutions based on the needs of the buyer.  

Additionally, the effectiveness of AI solutions is directly aligned to the data fed into the software. If inadequate quality data is used by AI technology, the solutions it will provide are unlikely to be of any real benefit. This report delves into numerous challenges the transport sector is facing, exploring themes of decarbonisation, solution procurement, organisational change and the quality of available data.

Data strategy must come before AI adoption 

In the realm of transport, AI is set to transform traditional practices. The adoption of AI promises to boost efficiency, safety and sustainability across the industry. But its efficacy is wholly reliant on the data upon which its algorithms operate. Inaccurate, incomplete or outdated data will almost certainly lead to flawed predictions, which reduce the positive impact of automated decision-making and undermine the objectives of AI implementation. 

Data quality, governance, and ethical considerations lay the groundwork for successful and impactful AI solutions. Rigorous validation and cleaning processes can identify any inaccuracies and ensure that data sets are diverse and representative of real-world scenarios. The continuous monitoring and updating of datasets to match current conditions will support tools in producing intelligent outcomes.

How AI can help decarbonise transport 

In 2022, the transport industry was responsible for 34% of the UK’s CO2 emissions, with the majority coming from road transport. AI stands to make a considerable contribution to reducing the impact of the industry on the environment.  

• Predict deterioration of infrastructure pre-issue 

The shift towards Electric Vehicles (EVs) places immense pressure on existing infrastructure. A predictive maintenance strategy leveraging AI can anticipate infrastructure deterioration, such as the upkeep of road surfaces to manage potholes, enabling proactive interventions and minimising disruptions.  

• Optimise transport plans 

AI solutions will enable the analysis of both historic and real-time traffic data to optimise routes for road vehicles, reducing congestion and therefore fuel consumption. Public transport planning can be improved with analysis of commuter journeys, allowing for more sustainable, people-centric urban mobility.   

• Increasing active transport adoption 

Leveraging AI algorithms to analyse user behaviour and preferences can facilitate the design of effective strategies for promoting walking, cycling, and other eco-friendly modes of transport. Despite the vast potential for AI to decarbonise the transport industry, it must be noted that the technology itself consumes vast amounts of energy. We must be aware of the environmental costs of storing and analysing huge datasets and include the impact of doing so in any decision-making in the future.   

• The role of change management  

AI will naturally bring change to an organisation; how this is managed will be the difference between the companies who succeed in this new era and those who do not. 

• Knowledge transfer 

Several speakers at the Transport AI event, across a range of positions in the sector, mentioned a distinct lack of in-house AI capabilities - driving the need for external specialists. These specialists need to ensure AI is not just implemented, but training and guidance is embedded within service delivery. Fostering a continuous learning environment can drive the AI industry forwards, minimising the risk of organisations falling behind.  

• Limited centralised governance 

At this stage, AI adoption is still heavily dependent on a limited number of specialists, meaning there is a lack of centralised governance for firms to rely on. But this creates the opportunity for innovative partnerships between public and private sector organisations to share talent, resources, and processes. 

• AI solutions will not solve all problems 

Throughout the conference, there was clear frustration with AI solutions trying to solve all problems, rather than focusing on specific client needs.  

• One size does not fit all 

AI is dominating the technology scene. Many startup companies are developing technologies in the transport space. However, the public sector is seeking solutions for niche scenarios, opposed to general problems, creating a reluctance to adopt ‘one size fits all’ AI technologies. This highlights the need for customer-driven AI development, ensuring suppliers work closely with clients in the transport sector to collaboratively solve niche problems with specific AI solutions. This approach would not only speed up AI adoption in the transport sector but allow AI SMEs (Subject Matter Experts) to grow and develop. 

• AI is not always the solution 

The craze to integrate AI has seen organisations forcing the use of AI tools to solve all business problems. Many speakers at the conference noted that AI is not necessarily the only solution to an organisation’s challenges and that all options should be explored. Having a clear understanding of the issues an organisation is facing is vital to the implementation of an effective solution.