· · · Institutional Repository

Reliability of travel times is an important indicator of the performance of a traffic system. The congestion caused by incidents is an important cause of the unreliability of travel times. Travel time reliability should be incorporated in social cost benefit analyses for infrastructure investments. With an accurate forecast of travel time reliability a well-educated decision of the consequences of infrastructure investments on travel time reliability, can be made. A large number of incidents (car accidents and breakdowns) are simulated in marginal traffic models MIC (marginal incident computation) and MaC (marginal computation). The research shows that it is possible to forecast travel time variability from door to door in case of an incident, with explicit simulation of incidents in a dynamic traffic model within reasonable calculation time. This could be done because of the usage of marginal traffic models, MIC and MaC, reducing the number of calculations needed and therefor the calculation time.

Travel and traffic information play an important role in traveller behaviour and traffic flows. Travellers have the possibility to use information on current and future traffic situations to make their travel decisions. Travel decisions by individual travellers subsequently influence the performance of a total traffic system. This makes behaviour of travellers an interesting topic to policy makers and road operators. In this thesis travel choice behaviour also plays a central role and we specifically take into account car drivers and en-route route choice. This graduation assignment was done at TNO and after an initial scope further focus was applied resulting in the goal of modelling driver traffic information use and subsequent enroute choice. After coupling en-route route choice to DRIPs as the channel for travel time information in this study it was decided to model in the ITS Modeller. Latter provides a micro simulation environment through which driver behaviour can be analyzed at the individual driver level. First opportunities for improvement in modelled driver use of information during en-route route choice were identified. The following research questions were used to further explore and implement a driver information model. RQ1 How can the DRIP information market be characterized and how do market dynamics affect driver route choice behaviour? Through a concise exploration the DRIP information market and visit to the Verkeerscentrale Midden-Nederland we made some interesting findings. These findings originate from the fact that the DRIP information chain is a complex system involving many technical systems but also a multitude of stakeholders. Some examples resulting from DRIP information dynamics are used to illustrate effects on route choice behaviour. The first example is formed by occasional manual operation of DRIPs by traffic controllers. During this process it appears information provisioning to drivers is prone to delays. Secondly it became clear that the small number of responsible parties for maintenance of DRIP control systems posses a large amount of knowledge seemingly necessary to maintain DRIP information quality. Both examples carry a risk of information quality deteriation and as such a the risk of losing trust of drivers in DRIP information. RQ2 Given the lack of a driver information evaluation mechanism in the ITS Modeller: How can driver route choice behaviour be modelled and implemented? This report identifies an apparent opportunity for improving driver route choice modelling under influence of DRIP travel time information. A driver information model is proposed in which car drivers exhibit updating behaviour on their travel time estimations and uncertainty about these times. The model fits in the current method of modelling route choice in the ITS Modeller. Our driver model is based on uncertainty about a drivers’ own mean travel time estimation and travel times provided by a DRIP. High uncertainty about his own estimation leads a car driver to update his expected travel time closer to that predicted by a DRIP. If the opposite is true then a driver updates his expected travel time closer to his initial estimation. In other words: “Updated quality perceptions are a weighted average of prior beliefs and observed quality. Weights reflect perceived reliability of prior beliefs and observations, respectively: when the traveller distrusts (trusts) his own observations, updated perceptions of quality are relatively close to initially anticipated quality (observed quality)” (Chorus & Dellaert, Forthcoming) RQ3 What are the route choice effects of the proposed driver informing model? The simulation phase illustrates the difference in driver travel time updating for different levels of uncertainty while all other simulation factors are kept constant. The implemented driver model indeed leads to different updated route cost under influence of travel time information. In the particular scenario comparison overall network effects due to difference in route choice behaviour were minimal though because of the small absolute difference in updated route costs. Further face validity analysis indicates that this small change in route choice behaviour early in the peak period appears to have a positive effect on travel time dispersion. This conclusion is not supported by significance and as such is made with reservation. Based on this study we recommend the following for further research: Practical recommendations: • Inclusion of truck drivers in use of the driver information model and future research. In this study we assume the group of drivers only consist of car drivers. In practise this is not the case and this study indicated that truck drivers show different route choice behaviour in light of traffic information, specifically route and travel time information through a DRIP. The current study overestimates the effect of provided information. • The utility function in our study consists of travel time and travel time uncertainty. In practise more factors may be important in a drivers’ en-route route choice consideration. Further research should investigate which factors are important. Practical recommendations: • During implementation of the model it appeared that an implementation at individual driver level would lead to too high computation times. We consider some opportunities for better implementation to be feasible. TNO has the opportunity to use server based computation. Running the improved driver information model in such an environment might result in acceptable analysis times.An other option is further evolution of the way in which the driver information model is applied to drivers. An implementation could be conducted analogous to current implementation of for example lane changing and car following models. • We assume the lack of influence of a DRIP on network familiarity. In short route information of a DRIP does not lead to inclusion of a route in a drivers’ route set. Further improvement of the driver information model ought to take inclusion into account. • We saw that in certain situations DRIP information delays occur outside of control of the CDMS (Centraal DRIP Management Systeem). It is deemed beneficial if the source of this delay, manual operation, is taken into consideration and how this affects DRIP information delay. Gained insights could be used to have a future calibrated driver information model more robust.

The increased transportation air movements in the past years has raised questions at the department of Airside Operations of Amsterdam Airport Schiphol (AAS) regarding the safety, robustness, reliability and utilization of the airside traffic system. To get more insight to this future situation, the current situation is analyzed first. Rather limited information about the current traffic system was known, besides the fact it consists of the main artery the ‘Rinse-Hofstra road’ and the connecting roads towards the airplane stands. Another observation within the area of AAS is the fact that the research is performed in a complex multi-level actor environment, involving many different stakeholders. Also the variety in vehicles on the airside is very large. The four aforementioned aspects are analyzed in detail and resulted into different improvement possibility proposals. Regarding the safety aspect, several specific analyses are performed; using the black spot methodology, all unilateral and two-sided collisions of the past 8 years are analyzed. This methodology indicated that the main problem areas regarding safety are located at the beginning of the piers. Another analysis estimated an average damage costs of €1612.99 euro per collision, which indicated the fact that a reduction of collisions will directly result in (large) financial savings. Finally, it appears that ‘human errors’ are the most frequent causes for collisions. The second aspect of which the traffic system is analyzed is robustness: “the ability to fulfill the function of which the (traffic) system is designed for, even in non-regular situations which differ (strongly) from regular user conditions“ (Snelder, Immers, & Wilmink, 2004). It appears, using this definition, that the viaducts crossing the A4 highway, the tunnels and the beginning of the piers are not very robust for several reasons; this aspect resulted in mostly infrastructural related improvement possibilities, to make it possible to overtake other vehicles more easily and being able to reroute traffic more easily in case of an accident. The reliability of travel times is the third aspect the traffic system is assessed on, by performing travel time measurements with and without viaducts on the routes. It appears that the F-pier could be reached in the same travel time using the viaducts, coming from the tunnel, although these measurements result in slightly more constant travel times using the ‘viaduct-route’. With the additional proposed improvement possibility of increasing the speed on the viaducts from 30 to 50 km/h, the F-pier can be reached faster by the ‘viaducts-route’ and now the E-pier can be reached in equal time by both routes. Increasing the speed on the viaducts will result in an increased use of the viaducts, which will lead to an increased utilization; this is the fourth aspect in this research. To get more insight in the number of vehicles using the viaducts and other strategic locations within the system, traffic volume measurements are performed by placing three cameras on the airside. It appears that an increased traffic volume of almost 40% of the viaducts can be realized in this way, and so reduction of the same amount on the other route, resulting in an improved utilization. Concluding, a total of 15 different improvement possibilities, with corresponding implementation strategies, are proposed within this research, of which 5 improvement possibilities are resulted as the best improvement possibilities regarding costs, effectiveness (regarding the four aspects) and stakeholder support. These four improvement possibilities are: rerouting traffic over the viaducts to maximize the utilization of the network, increase the cooperation between stakeholders (mostly Airside Support and Airport Authority), add an extra (safety related) training component for new users, remove the current green pedestrian crossings and reduce the number of traffic signs to create a more clear overview.

What Holland needs is a paradigm shift. Were at least four thousand kilometres of traffic lane short of even beginning to cope with congestion. While the government has been working on plans to sink more money into extra rush-hour lanes and the occasional stretch of tarmac, a team of researchers has been drawing a picture of the economically ideal road network for the country. Their findings are pretty shocking: We have simply been thinking too small by several orders of magnitude.

There is a growing awareness that road networks, especially in major urban areas, are becoming more and more vulnerable to unforeseen disturbances like incidents because of the fact that the level of congestion keeps growing. Therefore, robustness measures have to be taken. In this paper the robust road network design problem is presented as well as a solution algorithm for solving the problem. The solution algorithm combines expert knowledge with advanced modeling techniques. In this way the method can be applied to large scale networks. This paper shows the quality of the algorithm for a test network.

Much of the delay in transport networks is caused by incidents. Many indicators are developed to determine vulnerable parts of a network without simulating the network flows with an incident on each of the links. This paper lists indicators proposed in literature and cross compares them. Their values for all links on three networks of different sizes are computed. Among others, the order and the cross correlation of the indicators is compared. For one network the effects are also fully computed, running one simulation per blocked link. Different vulnerability indicators rank the links differently. None of the indicators produces a result similar to the full computation. We conclude that the listed indicators are complementary.

It is computationally expensive to find out where vulnerable parts in a network are. In literature a variety of methods were introduced that use relatively simple selection criteria (measured in real-life or calculated in a traffic simulator) to pre-determine the seriousness of the delays caused by the blocking of that link and thereafter perform a more detailed analysis. This paper reviews the selection criteria proposed in the literature and assesses the quality of these criteria. Furthermore, a multi linear fit of the criteria is made to find a better, combined, criterion to rank the links according to their vulnerability. The paper shows that different criteria indicate different links to be vulnerable. Also combined they cannot well predict the vulnerability of a link. Therefore, it is concluded that to find vulnerable links, one has to look further than link-based indicators.

The Dutch road network is, like many other road networks in the world, congested in the morning and evening peaks. The locations of congestion are quite often the same; this makes it relatively easy to take the delay of this regular congestion into account when planning a trip. However, as a result of disturbances, also unexpectedly large delays occur. If no measures are taken, the Dutch road network, especially in major urban areas, is becoming more and more vulnerable to disturbances like incidents. This PhD study proposes a framework for robustness analysis that includes definitions, indicators and a set of measures that can be applied to make the road network robust against incidents. Furthermore, a method is developed by which robust road networks can be designed given these measures. The method combines expert knowledge with advanced modelling techniques. The quality of the method is proven by applying it to a small test network. Finally, this thesis shows how the method can be applied to a large realistic network of Amsterdam and surroundings. The practical value of the research appears from the fact that parts of it have already been used in projects for the ANWB, Verkeer en Waterstaat, de Raad voor Verkeer en Waterstaat, DVS, de stadregio Amsterdam en Bart Egeter Advies. This research is supported by the TU Delft, TNO, NGI, Transumo and TRAIL.

The Dutch national road network has been developed over several decades. In the past, roads were constructed according to the then current spatial and transportation planning philosophies. Because the existing road network is a result of a long process of successive developments, the question can be asked whether this network is the most appropriate from the current point of view, especially taking in consideration the current socio-economic structure of the Netherlands. To answer this question an optimization algorithm for designing road networks has been developed. With this algorithm the Dutch road network has been redesigned based on minimization of the travel and infrastructure costs and by taking into account the socio-economic structure of the Netherlands. A comparison between the existing network and the new design shows that the redesigned Dutch national road network has significantly lower total costs than the existing road network. It is found that the construction of less roads with more lanes on different locations leads to a reduction of the total travel time and the total vehicles kilometers traveled.

There is a growing awareness that road networks, are becoming more and more vulnerable to unforeseen disturbances like incidents and that measures need to be taken in order to make road networks more robust. In order to do this the following questions need to be addressed: How is robustness defined? Against which disturbances should the network be made robust? Which factors determine the robustness of a road network? What is the relationship between robustness, travel times and travel time reliability? Which indicators can be used to quantify robustness? How can these indicators be computed? This paper addresses these questions by developing a consistent framework for robustness in which a definition, terms related to robustness, indicators and an evaluation method are included. By doing this, policy makers and transportation analyst are offered a framework to discuss issues that are related to road network robustness and vulnerability which goes beyond the disconnected definitions, indicators and evaluation methods used so far in literature. Furthermore, the evaluation method that is presented for evaluating the robustness of the road network against short term variations in supply (like incidents) contributes to the problem of designing robust road networks because it has a relatively short computation time and it takes spillback effects and alternative routes into account.

The road network in the Netherlands and in many other countries is becoming more and more vulnerable. Small disturbances can cause major disruptions on large parts of the network. The costs of this vulnerability can add up to several billions of Euros in the future. In this paper we present a new network design methodology (architecture) which focuses on improving the robustness and therewith reducing the vulnerability of a road network. The architecture for robust network design consists of the following components: • a specification of the design standards, • a functional analysis of the road network, • a design process, integrating network design and (future) spatial plans • a test of the quality of the robust road network design. The architecture can be used to design a robust road network for any area. To demonstrate how the architecture can be deployed we have, in this paper, worked out the various steps in detail for the network of the area The Hague – Rotterdam in the Netherlands for the year 2020. A comparison of the new robust network design with the policy network of 2020 shows that in general the robust network performs better, both in the situation with and without disturbances. The total travel time is 2.3% lower, the total distance travelled is 0.8% lower and the average network speed is 1.6% higher. Furthermore, the average number of vehicle loss hours in the case of accidents is 30% lower. Therefore, it is concluded that the presented architecture for robust network design allows us to improve the robustness of a road network significantly.

Transport networks in major urban areas are becoming more and more vulnerable to unforeseen disturbances in transport networks, like incidents. For the near future, we expect an increasing number of incidents with a large impact due to the overall increase of the traffic load. In this paper the hypothesis is tested that, if no measures are taken, the impact of incidents increases in the future and, therefore, the vulnerability of the road network increases. It is shown that the current network of the area The Hague-Rotterdam in the Netherlands is already vulnerable. If the demand increases, the increase in total travel time is more than linear with the increase in demand in the situation without an incident. The impact of incidents also increases when the level of demand increases. This results in the overall conclusion that it is necessary to make the road network more robust.

In the past decade an increase in research regarding stochasticity and probability in traffic modelling has occurred. The realisation has grown that simple presumptions and basic stochastic elements are insufficient to give accurate modelling results in many cases. This paper puts forward a strong argument for the further development and application of probabilistic models and argues that a realisation must arise of the detrimental effects of blindly applying non-probabilistic models to traffic where probability is rife. This is performed by the demonstration that deterministic and simple stochastic models will, in many cases, produce substantially biased results where variability is present in traffic. Prior to this demonstration, recent developments in probabilistic modelling are discussed. While the case for probabilistic modelling is strong in theory, the application of such modelling approaches is only possible with sufficiently developed models. However there are still certain challenges to be addressed in probabilistic modelling before a widespread implementation is likely. Remaining challenges for probabilistic approaches are therefore discussed and it is shown that computational efficiency, correlations between variables, and data gathering and processing all remain difficulties that have yet to be fully overcome.

Verstoringen op de weg, zoals incidenten, kunnen tot onverwacht grote vertragingen van soms wel meer dan een uur per persoon leiden. Dit geeft aan dat het wegennet kwetsbaar (niet robuust) is. In 2009 heeft Rijkswaterstaat een begrippenkader en indicatoren voor robuustheid opgesteld (Snelder et al., 2009). Hierbij is robuustheid gedefinieerd als de mate waarin een wegsysteem zijn functie kan behouden bij verstoringen, opdat er voor de weggebruiker geen onverwacht groot reistijdverlies optreedt. Als indicator is de extra reistijd als gevolg van verstoringen gekozen. Dit paper gaat in op de vraag hoe deze indicator kan worden geoperationaliseerd. Operationalisering betekent enerzijds dat de indicator uit data moet kunnen worden afgeleid en anderzijds dat de indicator met behulp van een model moet kunnen worden uitgerekend. De methode voor data-analyse staat centraal in dit paper. Hierbij is zowel naar de kans op incidenten gekeken als naar het gevolg. Een uitgebreide analyse van verschillende incidentdatabases geeft aan wat de kans is op verschillende soorten incidenten op verschillende wegen. Het blijkt dat wanneer gekeken wordt naar type wegvak, het incidentrisico voor een afrit het hoogst is, daarna volgen de oprit, weefvak/splitsing en het normale wegvak. Voor de normale wegvakken geldt in grote lijnen dat het incidentrisico afneemt naar gelang het aantal rijstroken toeneemt. Een kwaliteitstoets van de gevonden waarden geeft aan dat het noodzakelijk is om meer verklarende factoren voor het al dan niet optreden van een incident te vinden. Naast de kans op incidenten is het gevolg van belang. Incidenten hebben netwerkbrede effecten. Een incident kan tot file en dus vertraging op de weg zelf (1) en op andere wegen (2) leiden. Daarnaast wordt reistijdverlies geleden door mensen die besluiten om te rijden (3) en door extra files op de alternatieve routes (4). Tot slot kunnen kijkfiles optreden (5), zijn stroomafwaarts van het incident effecten te meten (6) en kan vraaguitval optreden (7). Dit paper beschrijft een methode waarmee op basis van twee nieuwe innovatieve data-analyse tools een nauwkeurige inschatting kan worden gemaakt van effect 1, 2, 5 en 6. Aan de hand van enkele voorbeelden wordt duidelijk gemaakt hoe de methode werkt. Bovendien zijn de geaggregeerde resultaten van een analyse van een deel van de incidenten die in 2009 plaatsvonden weergegeven. Hieruit blijkt onder andere dat het niet meenemen van fileterugslageffecten bij incidenten waarbij een maatregel is genomen (bijvoorbeeld het afkruisen van een rijstrook) kan leiden tot een onderschatting van 35% van het reistijdverlies.

Maatregelen die effect hebben op het verkeer zijn vaak ingrijpend, of hebben het doel ingrijpend te zijn. Daarnaast kennen dergelijke maatregelen vaak hoge kosten. Het kunnen inschatten van de verkeerseffecten van een maatregel is daarom erg belangrijk, en wereldwijd is er inmiddels een jarenlange ontwikkeling gaande in voorspellingsmethoden voor verkeerseffecten. Als de effecten van de maatregel goed voorspeld worden, kan gekozen worden voor de meest efficiënte oplossing. Het toepassen van prognosemodellen is een veel gebruikte methode om inzicht te krijgen in de effecten van maatregelen in het wegennet. Maar elk prognosemodel bevat zijn eigen representatie van de werkelijkheid, dus welk model kan nu het beste gebruikt worden? George Box verwoordde dit in 1979 als volgt: “All models are wrong, some are useful”. Bekend is dat statische en macroscopisch-dynamische verkeersmodellen verschillende resultaten berekenen in zwaarbelaste netwerken. Dit is bijvoorbeeld het gevolg van het feit dat statische modellen de file in de bottleneck plaatsen, in plaats van ervoor, zoals dat in de realiteit gebeurt. Het heeft echter ook te maken met de mate waarin de modellen rekening houden met terugslag van files. In dit paper wordt aangetoond dat de modelkeuze van invloed kan zijn op de informatie die voor beleidsbeslissingen (zoals de aanleg van nieuwe infrastructuur) beschikbaar is. Drie verschillende vormen van filemodellering binnen dynamische modellen zijn vergeleken: - een variant met verticale wachtrijen in het knelpunt, - een variant met horizontale wachtrijen voor het knelpunt, - een LTM-variant (LTM=Link Transmission Model) met eveneens een horizontale wachtrij voor het knelpunt, maar met een nauwkeurigere modellering van de op- en afbouw van files dan in de vorige variant. Alle drie de vormen van filemodellering zijn nauwkeuriger dan de filemodellering in de traditionele statische modellen. In dynamische modellen wordt bij alle drie de vormen van filemodellering de ontwikkeling in de tijd gemodelleerd. Bij statische modellen gebeurt dit niet. De LTM-variant geeft van deze drie vormen de opbouw en afbouw van files het meest realistisch weer. Dit blijkt onder andere uit een vergelijking die gemaakt is met een praktijksituatie. In een voorbeeld voor de regio Haaglanden is aangetoond dat de verschillende vormen van filemodellering andere filelocaties aanwijzen. Dit betekent dat de verschillende modeltypen ook andere investeringslocaties aanwijzen en dat de evaluatie van infrastructuurmaatregelen dus ook tot verschillende kosten-batensaldo’s zal leiden. Bij het nemen van infrastructurele maatregelen en bij het bepalen van de effecten hiervan moet dus goed nagedacht worden over de modelkeuze.

Onlangs heeft de Nederlandse media veel aandacht besteed aan een herontwerp van het Nederlandse wegennetwerk. Dit herontwerp is gemaakt om de behoefte aan infrastructuur te scheten. Bovendien kan een vergelijking met het bestaande netwerk inzichten geven in de punten waarop het netwerk verbeterd kan worden. Verschillende kranten en nieuwsrubrieken hebben over het herontwerp van het wegennetwerk geschreven. Waar in het begin de krantenartikelen nog volledig en correct waren, verdween bij de latere stukken al snel de nuance en werden ook de feiten niet altijd meer correct weergegeven. De vele reacties die daarop volgden van ministeries, verkeerskundigen milieuorganisaties en vooral particulieren waren heel divers. Hoe komt dit? Hoe erg is dit? Deze twee vragen worden beantwoord door enkele discussiepunten uit te lichten. Een voorbeeld van een punt waar veel misverstand over bestaat is de vraag of in het herontwerp nog file staat. Eén van de bladen kopte met “Nooit meer file” en in een andere nieuwsrubriek was te lezen “Van files komen we nimmer-te-nooit-niet-never-af”. Beide uitspraken waren gebaseerd op hetzelfde onderzoek. Het objectieve antwoord op deze vraag is dat in het herontwerp nog steeds files staan, omdat het financieel gezien niet optimaal is om het wegennetwerk te dimensioneren op een relatief korte periode van de dag waarop het heel druk is. Dit is een duidelijk voorbeeld van hoe feiten verkeerd kunnen worden weergegeven in de media. Misverstanden over de gebruikte terminologie en over het doel van het herontwerp zijn mogelijke oorzaken hiervan. Daarnaast beschrijven veel mensen hun eigen interpretatie van de resultaten van het onderzoek als de waarheid. Deze interpretatie wordt veelal niet getoetst bij de onderzoekers. Uit het feit dat de resultaten van het onderzoek regelmatig foutief worden weergegeven concluderen we dat expertise vereist is om de resultaten van het onderzoek goed te interpreteren. Het is vervelend dat artikelen niet altijd volledig en correct zijn, maar er zijn voldoende mogelijkheden om dit achteraf te corrigeren. Degenen die echt geïnteresseerd zijn, zullen uiteindelijk altijd contact opnemen met de onderzoekers en het is dan wel dankzij de media dat ze met het onderzoek in aanraking zijn gekomen.

Verkeersnetwerken in verstedelijkte gebieden zijn kwetsbaar voor onvoorspelbare gebeurtenissen zoals incidenten. Het komt nu bijvoorbeeld al regelmatig voor dat een klein incident tot file op meerdere wegen leidt en dat het uren duurt voordat deze file weer is opgelost. In de nabije toekomst zal naar verwachting de gemiddelde impact van incidenten groter worden, omdat de belasting op de weg groter wordt en daardoor de reservecapaciteit in het netwerk per tijdstap en over de tijd afneemt. Niets doen is dus geen optie. Verschillende maatregelen kunnen genomen worden om de robuustheid van het wegennetwerk te vergroten. Om het effect van deze maatregelen in te kunnen schatten is het van belang om robuustheid te kunnen kwantificeren. In dit paper is daarom een overzicht gegeven van verschillende in Nederland gebruikte definities, indicatoren en meetmethodes voor robuustheid. Uit het overzicht blijkt dat er nog geen eenduidige beslissing is genomen over welke definities, indicatoren en meetmethodes het best gebruikt kunnen worden. Het is zelfs de vraag of een keuze gemaakt kan worden tussen alle indicatoren, omdat ze immers allemaal een ander aspect van robuustheid laten zien. In praktijk zal echter toch altijd gekozen moeten worden. Bij de keuze van indicatoren en methodes is het van belang dat zoveel mogelijk aspecten van robuustheid beschouwd worden. Fileterugslag en alternatieve routes zijn de twee belangrijkste aspecten hiervan. Bij voorkeur moeten de indicatoren en methodes rekening kunnen houden met de ontwikkeling van files over de tijd in de situatie met en zonder verstoring. De mate van file en de locatie van de file zijn hierbij van belang. Dit betekent dat de te gebruiken methode met fileterugslag rekening moet kunnen houden. De aanwezigheid en het gebruik van alternatieve routes zijn het tweede belangrijke aspect. Tot op heden is het echter heel lastig om dit goed te modelleren, omdat weinig bekend is over het gebruik van alternatieve routes bij verstoringen. Het is daarom aan te bevelen om meerdere vormen van routekeuze toe te passen, om een zo goed mogelijk beeld te krijgen van de range waarin de resultaten kunnen liggen. In de toekomst zullen de methodes voor het kwantificeren van robuustheid verder uitgewerkt moeten worden om goed onderbouwde beslissingen te kunnen nemen over de te nemen maatregelen om de robuustheid van het wegennetwerk te vergroten.