Norway Moves Alongside Croatia and More as Preikestolen Plans Smart Forecasting to Ease Tourism Pressure

Image generated with Ai

Norway is using smart visitor forecasting in Preikestolen, one of its most popular destinations. Norway is joining the likes of Croatia and other European and Asian tourist destinations in employing new predictive tourism technology to help manage visitor flows and ease the burden on both the local environment and its residents. Around 400,000 hikers per year are estimated to visit the Pulpit Rock and the use of smart visitor forecasting is being trialed to anticipate the peaks and alleviate the congestion on the roadways, parking lots, trails, and bus routes that lead to the popular and busy viewpoint over Lysefjord.

The new approach is being explored through VisitorLAB, a pilot involving SINTEF and Lysefjorden Utvikling. Instead of relying only on historical visitor totals, existing information from several systems is being brought together and analysed to determine when tourism pressure is likely to rise.

The technology places Norway within a wider international shift towards predictive tourism management, with comparable approaches already being developed or used in destinations including Croatia, Japan, Italy, Slovenia, Finland and Spain.

Preikestolen Tourism Pressure Creates Demand for Smarter Visitor Management

Preikestolen has become one of Norway’s most recognisable natural tourism attractions. The dramatic rock plateau above Lysefjord draws hundreds of thousands of hikers, particularly during the warmer tourism season.

Approximately 386,934 visitors were recorded in 2023, compared with about 50,000 in 1992, illustrating the enormous long-term expansion of tourism at the site. Around 400,000 annual visitors can therefore be used as a reasonable contemporary estimate, although yearly totals can rise or fall.

The challenge is not simply the total number of people visiting during an entire year. Greater pressure can be created when thousands of visitors arrive during the same hours.

When that happens, additional demand can be placed on access roads, car parks, buses, toilets, hiking infrastructure and safety services. The walking trail itself can also experience increased environmental pressure, particularly where concentrated footfall causes erosion or damage to surrounding natural areas.

This has made the timing and distribution of visitors almost as important as the annual total.

VisitorLAB Uses Existing Data to Predict Tourism Crowds Before They Build

A different form of destination management is now being tested through VisitorLAB.

Rather than collecting an entirely new set of information, data that are already available are being combined. These include visitor counts, weather conditions, road traffic information and cruise arrivals.

A LightGBM machine-learning model has been used to examine patterns in Preikestolen trail-counter information. Data covering previous visitor activity have been combined with other variables to establish relationships between factors such as weather, seasonality, traffic conditions and tourism demand.

The system is intended to estimate visitor volumes approximately 24 hours in advance.

This could allow crowd management to become more preventive instead of reactive.

If unusually heavy visitor flows are expected the following day, destination operators could potentially prepare additional transport capacity, adjust staffing, organise traffic management or communicate expected busy periods before travellers begin their journeys.

The system remains a pilot project, rather than a fully automated real-time operating platform. Data still have to be obtained from several separate systems, while greater automation and additional datasets would be required before the approach could operate at a larger scale.

Even so, the project has demonstrated how information already being generated by tourism and transport systems can potentially be transformed into practical destination intelligence.

Roads, Parking and Cruise Arrivals Become Part of the Tourism Forecast

The importance of forecasting becomes clearer when the way visitors travel to Preikestolen is considered.

Private transport has traditionally played a major role in access to the attraction. Earlier transport analysis found that approximately 73% of visitors arrived by car, rental car or motorhome. Private buses accounted for another significant share, while cruise-related buses also contributed to traffic movements.

Large visitor peaks can therefore affect more than the walking trail.

Parking capacity may be tested. Roads can become busier. Shuttle services can experience sudden increases in demand. Groups arriving from cruise ships may overlap with independent tourists travelling by car or organised coach.

By combining road traffic and cruise-call information with hiking data, VisitorLAB is designed to provide a wider picture of tourism pressure.

The approach could eventually answer questions that simple visitor counters cannot answer alone.

Destination managers could identify whether a large volume of cruise passengers is expected to arrive during the same period as independent hikers. Bus operators could potentially prepare for heavier demand. Information services could encourage travellers to consider quieter visiting hours.

Visitor flows could therefore be distributed more evenly rather than being allowed to concentrate around the same limited periods.

Norway Moves Alongside Croatia, Japan and Italy in Predictive Tourism Management

Norway’s project forms part of a wider movement in which destinations are increasingly being managed through tourism data, artificial intelligence and predictive analytics.

Croatia, particularly Dubrovnik, provides one of the closest European comparisons. Visitor-counting technology, cameras, tourism information and machine-learning tools have been used to better understand and forecast visitor flows around the heavily visited historic destination.

Japan’s Kyoto has taken the concept directly to travellers through congestion forecasting. Smartphone location information, weather conditions, previous tourism patterns and monitoring technology have been used to help indicate when popular districts and attractions are likely to become busy.

In Italy, Venice has developed sophisticated visitor-monitoring infrastructure using people counters, mobility information and big-data systems. Predictive modelling has been incorporated into efforts to understand pedestrian movements through one of Europe’s most tourism-intensive historic cities.

The technologies are not identical. Different destinations use different data sources, algorithms and management structures. However, the central principle is increasingly similar: visitor pressure is being predicted before it becomes a problem.

Sustainable Tourism Requirements Give Forecasting Greater Importance

The growing use of visitor intelligence is also connected with Norway’s broader approach to sustainable tourism management.

Preikestolen is associated with the country’s National Tourist Trails framework, under which heavily visited natural routes are expected to be carefully managed.

Visitor numbers, trail conditions, environmental effects and safety requirements need to be monitored so that tourism does not undermine the landscape that attracts travellers in the first place.

The wider Lysefjord area is also connected with Norway’s Sustainable Destination framework. Under such programmes, tourism development is assessed not only through visitor growth but through environmental, social and economic effects.

Predictive data could strengthen that process.

Instead of examining impacts only after a tourism season has ended, pressure could increasingly be identified as it develops.

This may become particularly valuable in sensitive natural destinations, where excessive concentration of hikers can contribute to trail erosion, vegetation damage, congestion and increased infrastructure requirements.

Trail-management work has already been required around Preikestolen, demonstrating that rising tourism demand has physical consequences. Forecasting could therefore become another tool alongside trail improvements, public transport, visitor information and safety management.

Smarter Forecasting Could Change How Popular Natural Attractions Are Managed

The significance of the Preikestolen project extends beyond one Norwegian hiking attraction.

Many tourism destinations already know approximately how many people visit each year. The harder question is determining where those visitors will be tomorrow, at what time they will arrive and how infrastructure should be prepared.

Predictive systems are being designed to close that information gap.

If VisitorLAB is developed into a more automated platform, a shared dashboard could potentially provide tourism operators, transport organisations, destination managers and public authorities with a common picture of expected visitor movement across the Lysefjord region.

Decisions could then be based on expected demand rather than assumptions.

More buses could be scheduled when necessary. Parking pressure could be anticipated. Staff could be positioned before peak periods. Visitors could be advised about quieter times. Tourism businesses could also prepare for changing demand.

This represents a significant evolution in destination management.

Tourism growth would no longer be measured only through annual arrival records. Increasing attention would instead be placed on visitor distribution, infrastructure capacity, environmental protection and the quality of the traveller experience.

Norway’s Preikestolen pilot therefore reflects a much wider transformation already visible in Croatia, Japan, Italy and other destinations.

As popular attractions face increasing pressure from concentrated tourism demand, smart forecasting and predictive visitor management are being positioned as practical tools for reducing congestion while protecting destinations.

Norway is testing out ways to predict the number of tourists that will visit Preikestolen. The predictions can then be used to ease congestion and protect the popular attractions as well as the Lysefjord landscape from damage by too many visitors, and also to benefit the traffic conditions.

The test being conducted at Preikestolen is only just beginning, and the outcome is still uncertain, but the concept is simple enough. By predicting the number of people who are set to visit the area, it might also be possible to prevent traffic congestion and overcrowding by spreading the numbers out over a longer period than would otherwise be the case.