In Brazil, the foreign trade has been not used as pro-active factor in development strategy because, historically, the negotiations between the different participants of the export chain have presented conflicts. It is observed that each link intend to minimize its individual costs, which normally does not converge to the global optimum of the supply chain. Therefore, companies are being obliged to re-analyze its procedures, to use reengineering techniques and redefine the relations and models of its supply chains in order to reduce costs, increase efficiencies and gain competitive advantage.
To reduce such problems it has recently emerged the concept of Collaborative Transportation Management (CTM), in the new concept of collaborative logistics. It has been spread out from the year 2000 through Collaborative Planning, Forecasting and Replenishment (CPFR) approach, and CTM has been defined by experts as a helpful tool to provide reductions in the costs of transactions and risks, enhance the performance of service and capacity, as well as the achievement of a more dynamic supply chain (Silva et al., 2009).
As the exporter Brazilian companies are looking for higher competitiveness, they shall not act in an individual manner and start acting in a collaborative manner. Therefore, it is required a detailed sharing of data and information by the agents of the logistics chain, in order to compose a solid partnership. It is understood as agent each integrant of this chain, as in the maritime logistics chain: the producer company, road transportation, maritime transportation and the load to be transportated.
After bibliographic studies and contacting entrepreneurs of this area, it is verified that there is restrict scientific work exploring this subject comprising manufactory industries, freight contractors and maritime transportation, in order to contribute with exportation. Therefore, this study, which is part of a Ph.D. thesis in development, intends to summarily present an overview of the Brazilian exportation and its operation of manufactured exportation chain using maritime model, the definition of the new CTM logistic term, its implementation mode in a maritime logistics chain, as well as to begin the application of the concept of autonomous agents to model and analyze this transportation problem.
This is a bibliographic work and, regarding its nature, it can be considered qualitative, aiming at generating knowledge for further practical application in logistic problems.
The article is composed by 6 items, including this introduction, distributed as follows: the second item presents a short overview of the Brazilian exportation, the third item contextualizes the CTM approach and the fourth item presents the use of the concept of Autonomous Agents applied to logistic problems. The fifth item presents the prototype model of resolution of the maritime transport logistic and the sixth item presents the final considerations, as well as suggestions for the continuity of this study.
2. Exportation of manufactured products
The Economic Commission for Latin America and the Caribbean (CEPAL) classifies the exportable items in manufactured products, basic products and semi-elaborated in function, generally, of the used resources, work intensity, scale intensity and science applied to manufacture. Manufactured product is that resulting from a standard serial production process, by machines, tools and labor, currently also treated as industrialized product.
In order to contribute with the foreign trade study the National Institute of High Studies (INAE) recommends the need of comprising three critical matters:
a) To know the structure of global demand and the position of Brazil;
b) To define the concept of manufactured in the statistics of Government foreign trade and the goal to be reached vis-a-vis of global demand;
c) To identify the external barriers via protectionism, with global crisis or not, and internal barriers to the production and exportation to be overcome.
In the next topics it shall be presented the scenario of the Brazilian manufacturing sector, including applied volumes and figures, as well as other relevant factors such as barriers to the expansion of its exportation and the presentation of the mechanism when performing the exportation of products using maritime transportation (which is the focus of this work).
2.1 Global demand overview
In the year 2008, the global trades increased about 15%, where 60% of this percentage was concentrated in developed countries. Latin America has participated with 5.4% and Mercosul with only 1.7% of exportations. Regarding the sector distribution of the global exportations of the year of 2007, the latest statistic information available, prevail the products of higher aggregated value - machines, telecommunication equipment, transport, etc. – with a participation of 35.5% of the value. The Brazilian sales, however, represent only 0.7%.
According to the Ministry of Development, Industry and Commerce (MDIC) in 2008, 36.9% of the exportation value was referent to basic products – food and raw material in natura -, and 60.7% was referent to industrialized products. From this total, 46.8% were manufactured, including machines and equipment and consumer goods.
Besides the relevant contribution of manufactured products for the Brazilian economy, there are several problems providing the representativity of only 0.7% of the total amount of manufactured exported worldwide. One example to mention is the case of barriers imposed to the exportation chain, which may occur externally and internally. Regarding external barriers, it shall be highlighted: requirements of technical standardization and of technology nature, restrictions of environment nature, sanitary Standards and fitosanitary standards, voluntary agreements, genetic modifications direct subsidies to production via special financings, price or purchase warranties, fusions of companies or intercompany agreements, etc.
Regarding internal barriers, we can say those are resultant from a deformed vision regarding the importance of foreign trade of goods and services for the sustaining of economic growth, as a factor for the increase of job offers, induction of investments and motivation for the incorporation of new technologies. Upon the presence of this barriers, it is required a facilitating policy of the exportation mechanisms comprising the costs, routes and agents involved in the best possible manner.
2.2 Mechanism for the performance of maritime transportation in case of exportation
When deciding to export, the company shall observe the stages of the process, aiming at knowing its client market, its demands, habits and characteristics. It shall also define other points: the good transport mode, proper packing of the product in order to maintain its integrity, freight form to be adopted in the negotiation, as well as the company to perform the transportation, besides considering or not the support of an intermediate agent (freight forwarder) or NVOCC in the negotiation.
Considering that the most part of the exportations is performed through waterway mode and that this work intends to study waterway mode, it shall only discuss maritime exportation.
In maritime transportation the most important intervenient companies are the shipowner, maritime agency, NVOCC, load freight contractor, multimodal transport operator and cargo broker. Upon the performance of a sale or purchase it shall be established a delivery point of the goods, where the liabilities shall be shared between the vendor and the purchaser. These liabilities comprise costs and risks on the transaction, and the exporter and importer shall undertake them until the delivery point and from there on.
2.2.1 Transport negotiation
In case the alternative is sending goods as free general load, or even in pallets, the exporter or its agent (freight forwarder) shall be able to perform booking of a conventional ship of regular line (trough a NVOCC or not); or even freight a ship in case of a larger batch.
The next point is to decide through which port the goods will be shipped, as well as the terminal and shipowner to be used. Next, it shall be define the availability mean of the goods in the chosen terminal based on the cost of transportation, traffic time and deadline for the delivery of the goods. One possibility to be considered is the convenience to have a safety warehouse in another city, which means, to maintain a stock in some point to solve quick delivery problems or small quantities which cannot be always quickly produced or sent.
As a consequence of the planned logistics it is possible to reduce the costs of transportation, storage, loss of time on traveling and, finally, problems in the delivery and compliance with sales contract. Not mentioning demurrage, if established in contract, which is normal in freight contracting, due to loss of time of the ship caused by the shipping agent on shipping and landing operations. Therefore, in the decision regarding the port to be chosen by the exporter it shall be also considered the quantity of goods to be produced and shipped.
Regarding the negotiation rules it is important to mention that each country has its own set of rules for this kind of negotiation, but normally for internal use in its domestic market (Keedi, 2007). If each country intends to use its own rules in external operations, there shall be a large range in use, which can generate problems in the execution of foreign trade. Then, it was created a set of international uniform terms in order to facilitate trading, avoiding misunderstandings. This set was created by the International Chamber of Commerce (CCI), with headquarters in Paris. This set of foreign trade terms was named International Commercial Terms (INCOTERMS), which main objective is to settle conflicts emerging from the understanding of international contracts signed between exporters and importers regarding the transference of goods, the expenses arising from the transactions and liability for damages and losses.
2.3 How to operate in exportation? Individually or in collaboration?
After defining the general maritime exportation mechanism, it is observed that it is very comprehensive, comprising several variables and requiring updates regarding volumes, capacity of the ships, prices, maritime fees, fees, law, among others. One point to be discussed is how the exporter companies involved in this mechanism shall act: individually or in collaboration? If isolated, each exporter company shall analyze the best manner to perform the transport of its goods, analyzing the existent navigation companies, fees, acceptable volumes, transport mode of the goods and the proper port, as well as preparing the required documentation.
One possibility to begin collaboration is searching for help from a professional prepared to perform such activities, the freight forwarder to undertake, if possible, the joining of loads of other clients in order to achieve scale savings. It can be also performed directly through partnerships among exporter companies, with no contracting of this third party. Regarding shipowners, the collaborative transport can be also performed if a certain shipowner does not have sufficient volume to fulfill a ship, joining with others and sharing expenses and gains.
Such alliances are under constant transformation, altering its members to adjust to market instabilities or to expand services; those can look like instable associations in a first moment, but are actually flexible and eliminate more fragile competitors, expand the service coverage, enhance efficiency, punctuality and speed. Through this collaboration in the transport it is observed the formation of global logistics networks articulating productive areas, not only expanding service coverage but also frequency, efficiency and circulation rate of the goods (Souza, 2009).
The following item shall approach CTM, in order to explore this new logistics concept and trying to apply it in logistics chain of exportation of manufactured products through waterway mode.
3. CTM contextualization
Considering that the Brazilian GDP was correspondent to US$ 1.99 trillion in 2008 and that the percentage applied to logistics is of 12%, US$ 238.8 billion were invested in logistics costs this year. It can be stated that US$ 143.28 billion were addressed to costs with transports, which means, about 7.2% of the domestic GDP, a value much higher than the value applied in developed countries such USA, applying only 4.8% of the GDP in transportation. It shows that the transportation is a significant item in final accountability of resources spent on Brazilian products, either for the domestic market as for products to be exported, composing “Brazil Cost” (Tacla, 2003).
One manner to reduce the existing problems in logistics is trough collaboration applied to transportation between the several agents of a logistics chain, adding value to the whole process. It is also valid to mention that, for more benefits, the CTM shall act in the interaction of several logistics chains, forming the concept of logistics network. For Tacla (2003) the collaboration is performed through sharing of resources, mainly in the use of the same transportation equipment, contributing for the reduction of costs by enhancing the productivity of the transportation equipment. Therefore, the beginning of the Collaborative Transportation is the preparation of a shipping prevision, including the generation of orders and loads and, finally, the performance of the delivery and payment to the carrier (Silva et al., 2009).
For the implementation of the CTM it is required, therefore, systems allowing inter-organizational collaboration in a cost-effective and technologically compatible manner. Without such systems, the companies trying to implement the systematic may find invalid efforts which are difficult to manage.
3.1 Collaborative transportation operational system models
The general process of decision taking begins with the finding of the decision taking problem, which means, a challenge for future business operations such as long term operations in the transportation market when entering collaborative system (Bloos e Kopfer, 2009). Factors such as the increase in competition appear to be the main decisive factor and the entrepreneurs are looking at the collaboration as a manner to achieve geographic expansion.
According to Bloos e Kopfer (2009) entering in collaboration operational system in transportation implies in the exchange of client data, which can be a very sensitive operation. The process begins with the searching for alternative solutions and an accurate specification shall be performed on how the goal can be achieved and measured. After reaching, the costs and benefits are distributed during the transactions.
Besides the short literature on this subject, some works were found specifying operational systems for collaborative transportation, which may serve as benchmarking for the companies. For Krajewska e Kopfer (2006) the collaboration process can be described as a three-stage process: specification of individual costs of each collaborator, mapping of partner orders for possible exchanges between partners and attribution of the benefit offered by each participant for the collaboration.
Gomber et al. (1997) present a collaboration model for planning of transportation through the usage of freight agencies with several autonomous rewarding centers actuating in the achievement of orders and negotiation of prices for the execution of the orders of the clients. The works of Kopfer e Pankratz (1999) Schönberger (2005), Bloos e Kopfer (2009) and Grünig e Kuhn (2005) follow a similar approach.
Regarding the applied part, involving the construction of mathematic model and application, one of the first Brazilian works referent to the collaborative transportation subject is of Carnieri et al. (1983). It analyses the feasibility of the implementation of a representative central of the cooperatives in the State of Paraná (Brazil), aiming at minimizing the cost of soy transportation. The problem is modeled as Integer Programming and, due to its simplicity; the model does not comprise important restriction such as deadlines, routes and capacity of transportation, among others.
Tacla (2003) develops a methodology for collaborative transportation of soy and fertilizer introducing flexibility to the restrictions imposed by time windows and minimizing freight expenses. Using mathematics model and CPLEX solver of GAMS 2.50® and, for the operational level, develops a heuristic solved in Excel® suggesting Linear Programming problem. The work of Giesen et al. (2007) has as its scenario a heterogeneous fleet of ships transporting multiple products, from a set of source ports to a set of destination ports. The problem is formulated as Mixed Integer Programming (MIP) allowing establishing the route and e programming of ships, as well as allocation of products to the compartments of the ships. Due to the fact that the proposed MIP cannot be solved on time, a heuristic is proposed.
Different from the traditional approach, Novaes et al. (2009) consider in its analysis a dynamic environment for urban routing problem subject to traffic jams. Using the concepts of Sequential Analysis, the study comprises a warehouse and a homogeneous fleet of vehicles serving a certain urban region. Each vehicle is designated for a district, leaving the warehouse in the morning, performing the designated orders (tasks), returning in the end of the day to the warehouse. From a dynamic point of view, part of the planned service is addressed to other agents (vehicles) when the system judges the need, avoiding that the tasks are not performed.
As observed, there are different approaches for the transportation problem, involving several sources, destinations, companies and products. A possible manner to model the maritime transportation problem for the Brazilian exportation of manufactured products is by the recent application of Autonomous Agents. In the following item it is presented the concept of Autonomous Agents applied to logistic problems.
4. Application of the Autonomous Agents concept in the resolution of transport logistics problems
The most part of data process systems and logistics systems comprise several components, treated as agents, which usually actuates in an autonomous manner, which can also communicate and interact with the others, forming a network. Therefore, the Autonomous Agents form a community in a common environment where they actuate and transform based on rules, in order to reach their individual objectives. Besides being a new approach to logistics, several studies have been developed by the Collaborative Research Centre 637 Autonomous Cooperating Logistic Processes group in Bremen, Germany, with the purpose of detecting when and how the autonomous control can present more advantages than the classic control, especially regarding time, costs and robustness.
4.1 Autonomous agent concept
According to Hölscher et al., 2007 each autonomous agent has a goal, rules and control conditions and manage, in an autonomous manner, the application of rules in order to complete the established objective. The fact that the agents are based on rules provides basis for the dynamic of processes.
One direct explanation of its use in the logistics problem is handling each container, truck, company and ship as an autonomous agent, with rules to be observed in order to reach its individual goals (for example: the goal of the container is leaving from company A and arriving at destination B within Z days). With such information, the agents analyze, in an autonomous manner, the possibilities of actuation and take their decisions. Such feature allows the creation of a decentralized system, but still with autonomy.
The transport network is a graphic where the nodes represent locations; which means, warehouses, where the packets (goods) can be collected or delivered. The edges represent the connections between the warehouses. Each edge is labeled with the time required for travel through the connection represented by the edge. Trucks and packets are modeled as autonomous agents. Instances of these Autonomous Agents are represented as special nodes with single identifiers. In the example presented the transportation network contains two trucks (1, 2) and one packet (1). The truck nodes are labeled with a number representing the time the truck can be moving. In this case, truck 1 is allowed to move for eight hours, and truck 2 is able to move for sixteen hours (it can be equipped with two drivers). Both nodes of the trucks are connected to rectangular node which is labeled with the number and exclamation mark. The number defines the load capacity of the truck, in this case specified in tons. Truck 1 has a capacity of 6 tons, and truck 2 is able to load up to 12 tons.
The exclamation mark indicates the node of tour in course. The packet node is labeled with a number that specifies its weight. It is also connected to rectangular node of the tour, which is connected to the warehouse currently retaining the packet. The exclamation mark of the tour node of the truck indicated the node of the packet tour, and the edge labeled as ‘dest’ connects the packet node with its destination warehouse, which means, the warehouse to which the packet shall be delivered.
It is presented in the following item a model prototype for the resolution of the transportation problem of this work, considering the agents as autonomous actuating in a dynamic environment.
5. Prototype model of resolution of the maritime transportation logistic problem
Due to the fact that the matter of the study is very comprehensive, with different variables and agents involved, it shall not be handled as an enhancement problem.
Therefore, the proposed model has the objective of establishing scenarios constituting formation of load (composition of container) by the exporter companies and the composition of containers x ships, as well as navigation routes, considering dates, deadlines, costs of transportation, port fees on load moving, etc., in order to reduce the cost of the collaborative chain, or the time, if it is under its interest.
Such information is achieve in function of the load volumes moved, cost of location of a container, value of land freight until the base port to perform such moving, values of port fees and maritime freight, as well as in function of individual margin costs of the companies when allowing delays or advancements in the delivery of its goods to the receivers.
In the development of the model the behavior of the environment shall be judged as dynamic, with random transportation demands and requiring the allocation of the manufactured products in containers. Those shall be transported until the origin port to be arranged in ships dislocating in space during time, completing the maritime transportation. Both load (container) and the transportation mean (ship) shall be judged as autonomous agents, taking decisions on the performance of the best route between the origin of the load and its final destination, observing the restrictions imposed in order to achieve a feasible solution for the logistic network.
It shall be observed the possibility of existence of non-critical factors in the process, which shall observe the distribution of probability of occurrence, interfering in the decisions.
Another matter to be highlighted is the set of advantages offered to the participants of the collaborative chain: in order to achieve alterations in the shipping and delivery dates to its clients, it shall achieve gains with the reduction of freight resulting from the joining of loads; which means, the costs of the collaboration shall be lower than the individual actuation costs. Therefore, the proposed model shall comprise a systematic of distribution of these gains in order to make attractive the participation of the entrepreneurs.
5.1 Model considerations
Due to the complexity inherent to the factors involved in the performance stages of the maritime transportation, it is required to adopt some considerations for the modeling of the problem:
a) Land transportation of large load volumes, allocated in 20 feet containers, supporting a maximum of 24 tons.
b) Exporter companies can use the containers individually or in collaboration with other companies; therefore, the load can be closed or not.
c) Maritime transportation of large volumes in containers. Ships are not allowed to navigated empty.
d) Random transportation demand.
e) Dynamic actuation of the containers with the possibility of their storage in ports of the route used between the origin and destination of the load.
f) Dynamic actuation of the ships, altering its navigation routes or not, passing or not through pre-established ports.
5.2 Model description
The search for a feasible solution for the transportation problem shall also comprise the resolution of the vehicle routing problem, regarding trucks or ships.
It can be established a description of the model initially through the overview of the process presented below in six stages:
a) The companies, in an independent and dissociated manner, executing its respective business plans comprise the sale of its manufactured products. From this plan the logistics planning areas shall create a storage and transportation plan. In this stage, the proposed model interacts with the individual plans creating a collaborative transportation planning, initially until the source port, considering shipping and delivery dates in the destination, as well the definition of the volumes. The collaborative transportation is initiated in this stage.
b) To perform the truck routing between the collaborative companies until the source port.
c) Verify the possibility of prompt shipping in the ship or waiting in the port apron.
d) The ship verifies if there is load for shipping or landing in the ports of its route and chooses to berth in it or not.
e) After shipping, the container can follow in the same ship directly to the destination port or choose for the landing in some port of the route to achieve better arrival date in the destination or better cost, considering the possibility of advance or prorogation in the delivery of the goods. The collaboration in this stage reaches not only the operational planning levels, but also the areas directly linked to the business. The Sales departments of the exporter companies shall agree in advancing or delaying deliveries to its clients, opting for making more flexible or not the time windows.
f) The sequence of the model shall close the mechanism proposing an enhanced feasible set of the conjugated routes. The transportation programming shall be the performance of the transportation of the exportation orders, aiming at the minimization of the costs and compliance with dates, closing the cycle.
With the result of the application of the model it is expected an effective collaboration between the companies, with scale savings transferred to collaborators in order to enhance the efficiency of the logistics and the financial result of the companies, actually proposing the collaborative transportation approach.
When analyzing several scenarios, it shall be considered the reduction of the freight account of the whole system. It is possible when allowing the conjugation of loads between the companies when composing the containers, at the same time when using the same ground transportation until positioning the load in the source port, obtaining enhanced negotiation status and, consequently, scale savings. It is also possible to achieve the reduction of the freight with maritime transportation, looking for the best cost-effective routes, negotiating best prices with the maritime transportation company and allowing alterations in the dates of shipping and landing of the load in the final destination.
Most part of such date alteration, especially deliveries, can be performed without any additional cost, once the usual service level does not comprise strict time windows for the delivery of the loads, mainly regarding exportation where there are several variables in the process.
The conjugation of load can be judged as a simple event in a first moment, but actually the combination of several possibilities between the collaborators of the system makes its execution very complex. Therefore, to achieve success in the resolution of this collaborative transportation problem, it is critical the implementation of an structured mathematical model in order that the system presents a feasible and accurate solution, with a total final cost lower than the cost obtained with the routes composed by the individual companies.
6. Final Consideration
Among the several links existent in the composition of the logistics chain, in can be observed that the transportation activity is one of the most significant of the chain, once it is through transportation that it is possible to synchronize the availability of raw material on time, in full in the manufacturing units and the final products in the distribution centers further to the final client (Silva, 2008). In this context, therefore, the CTM has been used in order to allow the synchronization of the logistics network.
Therefore, considering the relevance of this tool and there is demand for the maritime transportation service by the exporter companies, it shall require researching and modeling the transportation problem discussed hereby, analyzing the possible scenarios which best understand the needs of the companies, in order to search the equilibrium between its gains and losses. As presented hereby, the Autonomous Agents approach applied to this type of problem may constitute a tool to analyze the actuation scenarios of the exporter companies requiring discharging its production through waterway mode.
Considering the lacking of studies in this area it is expected that this study may be useful as incipient step in the development of a supporting tool for the analysis and decision takings by the entrepreneurs. Therefore, for the continuity of the study it is required to calculate the model on a mathematical basis, develop decision criteria to distribute the costs and benefits among the main agents involved in the collaborative transportation and test them. From the calculated model it is recommended to implement a computationl code for the resolution of the transportation problem based on the Autonomous Agents theory interacting in dynamic systems. After that, it is recommended to simulate operations in different operational conditions similar to the corporate practices in order to test the model and validate it.
1. Bloos, Melanie; Kopfer, Herbert. (2009). On the formation of operational transport collaboration systems. Annals of the II LogDynamics International Conference (LDIC): 329-338. Bremen, Alemanha, 17-21 de agosto.
2. Carnieri, Celso; Simiema, Hélio Hipólito; Mazzarotto, Marco André (1983). Programa Integrado de Transporte de Soja. Anais do XVI Congresso da SBPO, Florianópolis.
3. Giesen, Ricardo. Munõz, Juan Carlos. Silva, Mariela. Leva, Mabel. (2007). Método de solución al problema de ruteo e inventários de múltiples productos para uma flota heterogênea de naves. Actas del Congreso Chileno de Inginiería de Transportes.
4. Gomber, P; Schmidt, C; Winhardt, C. (1997). Elektronische Märkte für die dezentrale Transportplanung. Wirstschaftsinformatik, 39 (2):137-145.
5. Grünig, R; Kühn, R. (2005). Entscheidungsverfahren für komplexe Probleme. Ein heuristischer Ansatz. 2nd edition, Berlin, Heidelberg, New York: Springer.
6. Hölscher, Karsten; Klempien-Hinrichs, Renate; Knirsch, Peter; Kreowski, Hans-Jorg; Kuske, Sabine. (2007). Autonomous units: basic concepts and semantic foundation, In: Hülsmann, M.; Windt, Katja (ed.) Understanding autonomous cooperation and control in logistics, German: Springer.
7. Keedi, Samir. ( 2007). Logística de transporte internacional. 3a ed. São Paulo: Aduaneiras.
8. Kopfer, H; Pankratz G. (2009). Das groupage-problem koopererender Verkehrsträger. Oper. Res. Proc., 1998: 453-462, Springer, Berlin Heidelber New York, 1999.
9. Krajewska, Marta Anna; Kopfer, Herbert. (2006). Collaborating freight forwarding enterprises – request allocation and profit sharing. OR Spectrum 28:301-317.
10. Novaes, Antonio G. N; Frazzon, Enzo M; Burin, Paulo J. (2009). Dynamic vehicle routing in over congested urban areas. Annals of the II LogDynamics International Conference (LDIC): 103-112. Bremen, Alemanha, 17-21 de agosto.
11. Schönberger, J. (2005). Operational freight carrier planning, Berlin, Heidelberg, New York: Springer.
12. Silva, Vanina Macowski Durski. (2008) Um modelo heurístico para alocação de navios em berços. Dissertação de mestrado do Programa de Pós Graduação em Engenharia de Produção da Universidade Federal de Santa Catarina, Florianópolis.
13. Silva, Vanina M. D. Zago, Camila. A. Abreu, Leonor. F. Coelho, Antonio. S. Gonçalves, Mirian B. (2009). La influencia de la gestión del transporte colaborativo para la eficiencia de la cadena de suministros. Anales del XIV International Conference on Industrial Engineering and Operations Management (ICIEOM), Salvador-BA, Brasil.
14. Souza, Erika. (2009). Alianças, América Latina e sistema marítimo portuário mundial. Anais do 120 Encuentro de geografos de America Latina, 3-7 de abril, Montevideo, Uruguay.
15. Tacla, Douglas. (2003). Estudo de transporte colaborativo de cargas de grande volume, com aplicação em caso de soja e fertilizantes. Tese de Doutorado. Programa de Engenharia Naval e Oceânica, área de concentração: Transportes e Sistemas Logísticos, Escola Politécnica da Universidade de São Paulo.