Wikis

Modelling the Multimodal Traffic Situation in the Port of Hamburg (Concept)

Pilot 4: Modelling the Multimodal Traffic Situation in the Port of Hamburg (Concept)

Jörg Dubbert informed 9.3.2009:

A piece of news concerning the Port Pilot Hamburg:

Recently tried to speed up the Port Pilot development process. The difficulty with this is that I have to do with external parties which support me on a goodwill basis. They are - fortunately - really interested in the idea because it seems to be quite relevant. On the other hand, the idea is complex and needs considerable development work. The if this is done on a goodwill-basis, their working priorities are hard to influence but we make progress slowly.
Actually, we have at the moment two local key partners, the Hamburg Port Authority (HPA) and DAKOSY, the information service provider for the Port of Hamburg. HPA have announced that the will send first data samples next week. Also next week, I will have another meeting with DAKOSY in order to agree next steps. DAKOSY is harder to get because they work commercially, and they might not be willing to provide data for free.
Another problem is that the logistic process needs to be modelled for this application. This is a considerable effort which we only can tackle in a group. I agreed with René to have a meeting in April which shall kick-off the conceptual work on modelling the process of container handling.

----------------------------------------------

Purpose of the Pilot

The purpose of the pilot is providing a prediction tool for the generation and distribution of road traffic depending on the ship arrivals and departures. The problem consists of two parts:

• the determination of a correlation between ship arrivals and road traffic generated by the container terminals
• the distribution of the generated road traffic on the road network (building on the available VISUM approach).

There are two levels of pilot development. A first and slightly simpler problem is to provide a model for traffic planning. This model would then predict the road HGV traffic volumes based on planned ship arrivals/ departures and based on statistical data on road traffic. This approach would support road traffic planning purposes.

The second and more complex problem would be to make the process dynamic based on the real ship arrivals and the online traffic situation based on measured road traffic data. This approach would support traffic management purposes more.

The pilot could be planned within ROADIDEA. It is not feasible to implement and demonstrate the pilot within the time frame and resources of ROADIDEA. 

Idea and Motivation

Main Ports in Europe benefit considerably from a remarkable increase in the growth of global trade. Transport flows between East-Asia and Europe dominate international trade relations. The Port of Hamburg is no exception. In the recent years the container transport volumes have risen dramatically, and prognoses say that the volumes will again double in the coming eight years. If these prognoses are true, several main ports in Europe will face considerable problems handling the hinterland connection, and the hinterland connection will become the bottleneck in the performance and competitiveness of the ports. Innovative traffic planning and management measures can support the mitigation of this problem.

Hamburg will face challenges in rail, inland navigation and road transport. Although the main hinterland transport is carried out by rail, road transport plays an important role in handling local and regional container transport, e.g. in the so-called “Wet Triangle” made up by the port cities of Hamburg, Bremen and Bremerhaven. This calls for suitable solutions for road connections and for a closer look at the correlation between ship arrivals/departures and the development of road freight traffic. ROADIDEA could consider the road hinterland connection depending on the traffic demand generated by the activity of maritime transport. This has to be done by a closer look at the transport chain of containers from/to the ships including the ship arrivals and departures, the transhipment process, the container handling and the road traffic. The normal road traffic conditions need to be taken into account as well, because road container traffic naturally mixes with it.

The road network of the Port of Hamburg and in the wider surroundings of the port is characterised by a high share of Heavy Goods Vehicles (HGV) traffic. A part of this traffic is generated by the loading and unloading business of the container terminals and eventually by ship movements. If it is possible to predict the correlation between ship arrivals/departures and the generated HGV traffic on the roads, a better traffic prognosis could be made. This could be done in a first step by a model for road traffic planning. Depending on the availability of online data, this approach could be made dynamic for traffic management in a second step by involving measured online data.

The availability of predictions would allow road traffic planners to design the transport infrastructure according to the future demand, to consider alternative transport modes and schemes or to plan more intelligent technologies for traffic control and traffic information. The availability of online predictions of the impact of maritime transport may lead to planned short-term traffic management measures in order to be able to handle peaks in road freight transport.

Situation at the Site

In a modern main port like Hamburg, container transport is the dominating branch of the port business. So, road container transport will have an effect on the general traffic in the road network which provides access to the container transhipment terminals and the other port facilities which are involved in the container handling process (e.g. the customs offices, the container depots, the veterinary office and logistics companies).

Hamburg has currently five significant container terminals in the port. All of them generate road freight traffic. By far not all containers which are transported by the ships end up on the roads as road traffic. They are treated in different ways:

• Most containers are transported from and to the hinterland by road transport.
• A considerable part of containers is transhipped to another vessel.
• A considerable part goes by rail transport.
• There is also a need to transfer containers from one terminal to another. This is mainly done by road transport. However, alternatively also barges are used.
• A few containers go by inland waterways.

The decision on the transport mode is taken by a huge range of forwarding agents, and it is unlikely to identify all decision makers here. So the container terminals have no influence over the chosen transport mode. Furthermore, container moves from and to container depots which are done by road transport may be triggered.

The containers are all treated differently on the container terminals. The time spans for how long the containers stay on the terminals vary around a statistical curve. So, the prediction of the exact time when a single container will appear as a container moves on the road is not trivial. Also, procedures at the customs office and other locations in the port may influence the vehicle moves.

The Hamburg Port Authority (HPA) is responsible for planning und operating the road network in the port area. This means, HPA is responsible for the road construction and the traffic planning for all roads in the port area. It is in HPA`s interest to keep the traffic on the roads fluent and to provide access to the container terminals and all other port facilities.

For the future, HPA considers a complete re-organisation of the container hinterland transport on the roads, introducing an inland hub. For the justification and planning of these measures, it would be helpful to model the generation of container flows. HPA does yearly traffic counts and has a statistical overview of the general development of road traffic in the relevant network. Also, the share between normal road traffic and HGVs can be determined. The counts cover selected locations in the net for a single day of the year. The counting takes place regularly once a year. Currently, there is no online traffic data available.

The Hamburg Port Authority is currently planning the establishment of a port-related traffic management system. Part of this system will be a road traffic model for the port which is now built up step by step. An important feature of this model should be a prediction of road freight traffic depending on the maritime transport activity.

The Challenge for Innovation

The determination of the correlation between ship departures/arrivals and the generated road traffic is most likely the most demanding part of this investigation. The process is not so simple that a direct correlation between ship movements and the activity on the road network is obvious. For example, a ship arrival would not cause tangibly a larger amount of trucks coming to the port. The problem is more complex and the following factors have to be taken into consideration. More likely is a correlation where the overlapping processes of container handling at all terminals lead to an overall logistic situation which correlates with the overall road traffic situation. Factors to be considered may be:

• the fact that the Port of Hamburg has five container terminals working in parallel all generating road traffic,
• the fact that only a certain share of handled containers reappear as road traffic,
• the amount of containers loaded and unloaded by a ship,
• the fact that several ships are loaded and unloaded at the same time,
• the fact that the size of the ships varies,
• the fact that there are 20 ft and 40 ft containers,
• the final destination of the containers in the hinterland,
• the varying times for a container to stay in the terminal,
• road container traffic which is generated by the transfers between the terminals,
• road container traffic in relation to the container depots,
• opening times of the different organisations,
• generated trips with empty containers and chassis

Most likely, the solution cannot be found by tracing the moves of a single container. A statistical model needs to be generated which defines input and out terminal gates together with a distribution of their likely destinations/ origins. In a second step, these HGV moves must go into an O/D calculation, and it must be calculated how the generated HGV traffic flow is distributed in the road network and how it mixes with the general traffic flow which has nothing to do with road container traffic. Here, delays at customs gates and other service stations need to be considered as well. As a bonus, these processes could be made dynamic based on the online measured data such as

• real ship arrivals (interface to AIS system for ship tracking)
• measured road traffic data.

Available Data Sources

The following types of data (needs further clarification) are considered to be of use in the project:

• Traffic volumes in general in the Port
• Lorry volumes
• Ship arrivals and departures
• Container Handling Times
• Container Throughput of a Terminal
• Number of containers delivered
• Number of containers sent

Currently there are two identified suitable data sources:

• DAKOSY Datenkommunikationssystem AG is the transport information service provider for the Port of Hamburg. The company operates a number of applications (SHIPS, TRUCKSTATION, ZAPP) which contain interesting data sets about:
  • all planned ship arrival and departure times,
  • all container numbers of arriving and departing containers on board the ship together with the ship and the destination and transhipment terminal,
  • historical data about container truck movement at terminals
• The database of HPA contains historical traffic counts on all important road network parts. The HPA maintains a model for container traffic including ship moves, berth times and amounts of cargo to be loaded and unloaded, including simulations of ship moves.

These sources would need further detailed investigation. Furthermore, it needs to be investigated from where further suitable information can be obtained. Main data sources and the necessary background information may be found within the container terminals and the container transport companies. Online data in real time is not yet available.

Outcome of the pilot will be the number of lorries on the port-related road network in a snapshot situation depending on a certain predicted ship arrival-departure situation.  Assuming the system is running, there would be an offline simulation for planning purposes. In goes the expected ship arrivals, out comes the container lorry volumes in a certain period. The possible use case for the data is a simulation of the traffic loads depending on the ship arrivals/departures.

The components that possibly should be installed into the ROADIDEA platform are e.g. digital road maps, database, simulation software to be developed and an output representation.

Necessary Local Partners

The pilot needs the co-operation with HPA, DAKOSY Datenkommunikationssystem, HHLA (largest terminal operator) and maybe the container transport companies.

Necessary ROADIDEA Partners

The pilot needs workforce for data analysis and concept development. First of all, a model specialist needs to find a suitable mathematical model for the processes. The correlation between ship arrivals/departures and generation of road container traffic needs to be but into a mathematical set of formulae. Furthermore, the relationship with the existing modelling tools for the distribution of road traffic like VISUM needs to be defined.

• AMANOVA for support in modelling
• VTT for concept building and evaluation
• DLR for concept building

Next Steps and Outlook

The planning process needs to be started up together with HPA. HPA has declared an interest in this pilot under ROADIDEA. DAKOSY Datenkommunikationssystem has announced the willingness to give support with data.

Next steps should be:

• Setting up a local team consisting of ROADIDEA members and local partner and pilot kick-off
• Collection and Analysis of identified data sources with DAKOSY and HPA
• Clarification of conditions for the receipt of data
• Further conceptual ideas on the model and the modelling concept
• Identification of the need for further research into the process.

Estimated Pilot Schedule

The aim is to start the pilot during Q1 of the year 2009.