Mobility is something so essential to our daily lives that only when it’s absent we notice it. Traveling from city to city or between different destinations within a city is so commonplace we don’t even think about it. However, ineffective or poorly executed road maintenance, road safety inspections, and traffic management can cause restrictions on mobility and discomfort while traveling. In the most extreme cases, it can even put the travelers’ lives at risk. With the right approach and the help of meaningful data, the maintenance costs can be kept low, roads smooth, travelers safe, and traffic running.
But first things first. What do these terms actually mean and what’s data got to do with it?
Predictive road maintenance: why and how?
In road maintenance, three types of strategies can be distinguished: reactive, preventive, and predictive (the latter two are also known as proactive approaches). The reactive approach is the costliest strategy since the action is taken only when defects have occurred. This means that a road or asset items need extensive repairs or have to be replaced altogether. Preventive maintenance may sound like a good choice but looking at it from a financial perspective, it is not too effective. With this approach, the maintenance works are conducted in fixed intervals and some that could have been postponed are done when they are scheduled regardless of the state of the road or asset items. The best choice in terms of cost-effectiveness is predictive road maintenance. Predictive road maintenance is exactly as it sounds – the when and where a road or other asset items need repairs is predicted using various statistical methods.
The most valued predictive analytics method for road maintenance is using machine learning techniques (1). Although somewhat time-consuming, using these techniques is worth the investment. This is because the AI learns from the data, it’s not biased to any particular assumptions as humans are, and it doesn’t make any prior simplifications. This way, the AI can find patterns in the data that a human could have missed, making the predictive accuracy better than with other analytics methods (1).
Another positive aspect that comes with predictive maintenance is that during the data gathering process all the road defects and asset items are mapped. Moreover, with advanced technical solutions, such as EyeVi’s products, it is possible to avoid manual data gathering and insertion into the databases altogether, making the whole process a lot faster and cost-effective (2). In addition, using these solutions makes it possible to easily update the gathered data and use it in future analyses. Thus with predictive maintenance, the repairs are timely, the roads run smooth, and costs are optimized.
Road safety inspection: how does it work?
Similarly to road maintenance, the approaches for risk assessment on road safety can be reactive or proactive (3). The reactive approach is crash-based identification – locations with frequent crashes are pinpointed and further risks of accident occurrence are assessed. However, from the viewpoint of travelers, it would be better to prevent crashes from happening by using a proactive approach, i.e., the risks are assessed and analyzed based on data that presents various contributing factors rather than on data from crash reports. This means that accurate assessment of risk levels for a particular road can be calculated before accidents even happen. To compile data necessary for this kind of analysis, road safety inspections (4) are conducted. These inspections are done on already finished roads at different time points, such as during maintenance procedures or when a formal inspection has been ordered.
According to PIARC (4), road safety inspection consists of four steps: 1) desk study, 2) on-site field study, 3) road safety report, and 4) implementation of remedial measures. Important here is the second step since, in this step, the on-site data is gathered. In general, this is done by using various devices that measure and record the road furniture and asset items (e.g., traffic signs, road surface, and road barriers, but also the angle of curves and declines and slopes of the road). Based on the assessment of this data, the trained personnel can detect possible issues and take action.
Traffic management: a complex assignment for various systems
In addition to road maintenance and safety checks, an aspect that has a major role in the comfort and safety of the travelers is how smoothly the traffic is running. Cities can be highly complex in terms of how the traffic works and how much effort it takes to keep everything going. For smooth operating, large cities depend on traffic management systems (TMS). The main goal of these systems is to prevent traffic jams and keep the traffic running smoothly and effectively. These systems collect various dynamic traffic-related data from different sources, such as vehicles and traffic lights (5). While this data enables the systems to smooth the traffic flow by, for example, setting the time intervals of the traffic lights, static information is also needed.
This static data includes various constantly updated information about different asset items and street conditions. Such as the type, state, and location of traffic signs. Information about road attributes, such as its width, condition and location of road markings, the number of lanes a street has, curves, inclines and declines. This kind of data is necessary because, without it, the overview of the city’s infrastructure is incomplete.
Gather meaningful data and connect the dots
So, the dots that connect predictive road maintenance, road safety inspection, and traffic management are data points. Data points that entail the information that is needed for mapping the ongoing situation and gaining the best possible outcome. However, currently these data points are gathered separately, for a single purpose, and, more often than not, manually, e.g., defects data for road infrastructure maintenance. With this approach, the data gathering is costly and its use ineffective. But EyeVi is here to help you get this data in one go and make it usable for various purposes.
EyeVi has developed a solution that captures different sets of data, such as geodata, panorama, ortho, and pointcloud datasets. This data is then processed so that it becomes meaningful and usable for various purposes, e.g., in road assessment management systems, safety analysis, and traffic management. In addition, this solution makes it easy to keep the data constantly updated, since it is by far less time-consuming than doing it the good old-fashioned way (2). The time-effective data capture, AI-based data processing methods, and making the data to be usable for various purposes make EyeVi’s solution cost-effective and sustainable.
So, if you wish to take part in creating better future for your industry, connect the dots and use meaningful data.
- Karimzadeh A, Shoghli O. Predictive Analytics for Roadway Maintenance: A Review of Current Models, Challenges, and Opportunities. Civil Engineering Journal [Internet]. 2020 March [cited: 2021 June 14];6(3):602-625. Available from: https://www.civilejournal.org/index.php/cej/article/view/2005/pdf
- EyeVi Technologies. The use of artificial intelligence (AI) in defect detection of roads infrastructure [Internet]. 2021 May 10 [cited 2021 June 14]. Available from: https://eyevi.tech/news/the-use-of-artificial-intelligence-ai-in-defect-detection-of-roads-infrastructure/
- PIARC. Road Safety online manual. [Internet][cited 2021 June 14]. Chapter 10.1 Introduction. Available from: https://roadsafety.piarc.org/en/planning-design-operation-risks-issue-identification/introduction
- PIARC. Road Safety online manual. [Internet][cited 2021 June 14]. Chapter 10.4 Proactive identification. Available from: https://roadsafety.piarc.org/en/planning-design-operation-risks-issue-identification/proactive-identification
- de Souza AM, Brennand CARL, Yokomaya RS, Donato EA, Madeira ERM, Villas LA. Traffic management systems: A classification, overview, challenges, and future perspectives. International Journal of Distributed Sensor Networks [Internet]. 2017 [cited 2021 June 14];13(4):1-14. Available from: https://journals.sagepub.com/doi/full/10.1177/1550147716683612