Insights A better customer experience on the M25

Research was commissioned by Connect Plus and Connect Plus Services to understand road user perception of concrete road surface treatments, to help address customer pain points.

Atkins and Accent employed digital technology, advanced analytics, engineering, market research and visualisation tools to collate ‘in the moment’ customer experience data. Integrating audio and smartphone-generated road condition data with visual, telemetric, biometric and geospatial data for the first time, visual dashboards were created, providing a comprehensive overview of how road users perceive different road surface treatments.

Connect Plus (CP) and Connect Plus Services (CPS) sought to address this on the 30-year Design, Build, Finance and Operate M25 contract, one of Europe’s busiest motorways and key arterial link routes, by commissioning research to:

• Understand road user perception, behaviours and preferred treatments(s) for concrete road surface treatments, via qualitative and quantitative data.
• Create outputs to inform engineering decision-making for concrete road assets whilst considering customer requirements.

An innovative methodology was developed (enhancing approaches already used for HE), to collate “in the moment” customer experience data, for a true reflection of how the surface treatments would perform for users in the “real world”.

Dashcams, smartphone apps, fitness trackers, and pre-/post-journey interviews were used to monitor road user behaviour and responses. Qualitative and quantitative data (visual, audio, telemetric, biometric and geospatial) was also collated across six different surface treatments.

Atkins collaborated with specialist market research agency Accent to define the sample:

• Different driver types (business, commuter, leisure)
• Vehicle types (4-wheel drives, small/mid-sized, estate/large)
• Vehicle age (under 2, 2 to 5, more than 5 years)

19 independent participants were recruited to undertake a minimum of four “spontaneous” and “prompted” journeys to assess the effect on drivers and their vehicles across a designated trial area on the strategic road network.

By combining the range of datasets, a comprehensive analysis of road roughness, heart rate, video and in-car noise across the different concrete surface treatments was undertaken. Digital tools were carefully selected to enable this e.g. Feature Manipulation Engine (FME) (successfully used on other types of road projects), Racerender (introduced from the motorsport sector) and bespoke Python code (developed by the team for audio analysis).

Building on best practice developed by Atkins to analyse driver behaviour, the methodology was enhanced to focus on customer experience of assets for the first time.

Participants were provided with comprehensive briefing notes including visual instructions to reduce the risk of any user error. To minimise distraction and ensure a safe, natural driving experience, participants undertook journeys alone in their own vehicles, with the following data captured:

• Dash cams – provided internal and external footage to understand the context of the journey from a driver perspective along with visual confirmation of the precise location of the car on the network (i.e. section and lane) and any external contributing factors (i.e. other traffic).
• Audio – converted from the dash cam footage using machine learning, to analyse the changes in amplitude of the transitions between the different road surfaces in the trial
• Road roughness – captured via use of a smartphone app.
• Heart rate data – monitored via fitness tracker watches to identify points of stress or trends in physical response.
• Location and speed – captured and cross checked across the fitness trackers, dashcam meta data and app content.
• Post-journey interview – provided a qualitative account of the customer’s perception of the journey.

The dash cam footage was processed using software utilised in the motorsport industry to combine inward and outward facing footage, along with the audio output, roughness output from the app, the heart rate from the Garmin watch, and a map of the journey using geospatial information also collected by the watch. Combining all the factors allowed an in-depth analysis of the different variables and a spike analysis of the roughness, heart rate and changes in in-car noise. This highlighted correlations and trends across the data types against the different surface treatments. The output of this analysis was combined with the qualitative research to understand the road user perception from an overall perspective.

Analysis of the data captured from the in-car smartphone app and the audio converted from video footage was compared to the more traditional asset monitoring of engineering performance conducted over the previous 18 months. The in-car noise was validated during the sound analysis, as this aligned across each surface treatment with trends of external noise measurements (CPX) captured as part of the conventional asset monitoring work.

To ensure quality and consistency whilst managing significant volumes of data across multiple sources, formats and types, automated, repeatable and traceable processes for handling the data were developed and implemented.

Research objectives were also agreed, with progress updates shared via collaborative workshops with CPS, CP and HE. The visualisation tools for analysis strongly helped with dissemination of findings to key stakeholders.

The objective of the trial was achieved by understanding the impact of each treatment type and customer preferences, and the findings are now being used to inform a more customer-focused approach to managing concrete road assets, national policy decisions and operational interventions throughout the lifecycle of the M25 highway asset.

Other benefits of the innovative research have included:

• Enabling subtle changes to be identified which may have been missed through more traditional research methods.
• Providing a powerful tool to visualise analysis and communicate findings to stakeholders.
• Noise monitoring provides insight into environmental impacts.

Collating both qualitative and quantitative data including video, innovative in-car sound levels, speed, location, road roughness, heart rate and interviews, provided a comprehensive insight into road user perception of different road treatment types – enabling CPS, CP and HE to be better informed.