Engineering better roads through machine learning

Wenxin, Senior Pavement Engineer

In a connected and ever-changing world, pavement engineers like Wenxin are shaping the future of transport through innovative pavement design and sustainable asset management, making journeys easier, safer, and more reliable.

“I am a Senior Pavement Engineer in AtkinsRéalis Digital Asset Management team. I have worked as the pavement discipline lead in a number of National Highways’ major projects. In these projects, I led a team of people based in both the UK and India, to deliver the pavement design for motorways and principal roads.

I have also worked as an asset management consultant for National Highways, one of our biggest clients. By predicting the future deterioration of their pavement assets, we help them determine how best to use their budget to improve the performance of the whole road network for the road users.”

Everyone in the Digital Asset Management practice is constantly seeking out new ideas and adopting new technologies to keep improving capabilities for our clients. Wenxin has been part of the team since 2016 and has stayed with AtkinsRéalis because of the many opportunities for learning and development.

“When I look back on my time with AtkinsRéalis, I find that I’ve been able to pick up new skills every year. I’m currently studying machine learning and Python. Once I’ve learnt Python, I think it will be easier to go on to learn other computer languages.”

Wenxin started studying machine learning while working on projects for High Speed 2 (HS2) and East West Rail 2 (EWR2). By analysing the client’s pavement survey data, he was able to build an updated deterioration model with increased accuracy to help them reach an agreement regarding the use and maintenance of the local road network.

“To build the HS2 railway, HS2 Ltd needed to transport the construction material and equipment via local road networks. Therefore, we helped HS2 Ltd to assess the performance of construction routes, using deterioration models based on engineering rules and machine learning models. In the EWR project, we assessed the pavement construction routes prior to construction and proposed a maintenance plan accordingly.

Due to the large volume of data to analyse in these projects, I also developed coding using the Feature Manipulation Engine (FME) to automatically process the data with high efficiency.”

Wenxin’s approach to problem solving stems directly from the culture of the wider Digital Asset Management team. By respecting each other’s working practices, they make sure that there’s a psychologically safe environment in which to try new things – it’s fine to make a mistake, and it’s encouraged to shout out if you need some support.

“If you are a new engineer joining our team, you don’t need to be shy. We try to be kind with each other and support each other, even if we make a mistake. When I first joined AtkinsRéalis, my practice manager and line manager never criticised me – they only wanted to support me and my progress. And that’s how I treat the others in my team too.”

This attitude has been very successful for Wenxin’s team – in the last three years, they’ve won two internal recognition awards for their work in pavement design and have even been recognised for their contributions to the field by the Institute of Asset Management.

“I think those have been my proudest moments – standing on that stage and seeing the work I’ve done – that our team has done – being recognised for our contribution to the company. The latest award was for some cutting-edge research we did for the M25 Connect Plus service to use 50% recycled asphalt in new roads, to minimise the carbon impact of maintaining the M25. This project also won the Sustainability and Climate Impact Award at the 2021 Asset Management Excellence Awards. In addition, I have published a paper in the ICE Transportation Journal as a co-author based on the findings from this cutting-edge research.”

Join Wenxin in Engineering better. Discover how at InsideAtkinsRéalis.