Predictive Analytics: Maximizing visitors while preserving nature?  

The last years have shown that more and more people visit the Veluwe. The region is well known for its diversity: you can go for a peaceful hike, cultural activity or bike tour. The leisure possibilities are endless, tourism is booming. In fact, the region is becoming so popular that the tourism sector is changing its goal from attracting visitors to the Veluwe to managing crowdedness in the Veluwe region.

This development requires an active change in mindset for VisitVeluwe, the responsible party for organizing tourism in this region. Their goal is to spread visitors over different parts of the region, as well as spreading visitors more equally over time. This spreading of visitors has two main benefits. On the one hand, it reduces over-crowdedness. On the other hand, it maximizes visitors to the Veluwe without causing much damage to the flora and fauna. This second part should not be forgotten, since allowing many people to enjoy the Veluwe continues to be important for VisitVeluwe.

‍“There are places where the number of visitors has a negative effect on nature, the quality of life of residents and the experience of visitors. On the other hand, there are also places and times when that is not a problem. We want to get the balance right. To achieve this, it is important to spread the visitors in time and space. It helps enormously to be able to predict how busy certain area’s will be so that we can anticipate.” - Pim Nouwens, VisitVeluwe

‍However, managing the crowdedness of an area as large and diverse as the Veluwe is no easy task. VisitVeluwe has already taken several steps towards getting an insight into the crowdedness in the Veluwe region. A clear example is the already running Crowd Monitor, which helps visitors and managing parties understand in which places and at what times over-crowdedness is most critical.

Yet, the Crowd Monitor is still a reactive model, which will only get them so far in tackling the issue. The next step is being able to take preventive measures based on a pro-active model. In order to do so, they would need to understand where over-crowdedness will become an issue before it actually does. A pro-active model offers multiple benefits. Firstly, if VisitVeluwe have an understanding of the locations where crowdedness might occur in the near future, it can help them in managing the visitor streams towards other locations on the Veluwe. Next to that, a predictive crowdedness model can also help visitors in deciding on when to visit a certain place on the Veluwe. Not only will this assist in managing the crowdedness, but it might also improve the visitor experience, as most of us prefer a calm environment when visiting a place like the Veluwe. A final benefit of a predictive crowdedness model is that it can help tourist attractions on the Veluwe in understanding how many visitors they can expect and the consequences this brings for them (workforce planning, ticket prices, etc.).

In order to take these preventive actions, VisitVeluwe and Hammer developed a predictive model based on historical visitor data. The model takes visitor data as basic input and combines it with several predictor variables to ensure accurate location-specific predictions. These predictions can subsequently be used to create a crowdedness timeline, that will help to identify over-crowdedness issues. Due to the large variety of locations on and around the Veluwe, we opted to use a neural network model for this project. The architecture of such a model means that it is location independent and can be trained on any location with available historical visitor data. The nature of the predictive model will allow it to improve over time without much human interaction, aslong as it gathers real-time visitor data. This means that the current project lays a strong foundation for better crowd management solutions in the future and can be integrated into VisitVeluwe's already existing Crowd Monitor system.

‍“Together with Hammer, we investigated the possibilities. It was a new and quite complex issue, but I am very satisfied with the approach, skill and cooperation. It looks like we have developed a valuable tool.” - Pim Nouwens, VisitVeluwe

‍This predictive model leads to better preservation of the Veluwe, while at the same time allowing as many visitors as possible to enjoy the magnificent experiences the Veluwe region has to offer.

Source: Hammer, Market Intelligence