PROCESSING SYSTEM FOR THE ANALYSIS AND OPTIMIZATION OF TRANSPORT ROUTES

Abstract

The invention relates to a processing system (40) for analyzing transport routes, wherein this system is configured to receive data defining several routes (1, 1a) on a map of at least two dimensions, each route representing a transport. The system is also designed to receive data defining a set of nodes (10-27) on the map, each node (10-27) having a proximity zone (30). For each route (1, 1a), the system performs a mapping procedure to create a sequence of nodes comprising nodes (11, 15, 17, 19) traversed by the proximity zones (30) associated with each route (1, 1a).

Description

Title of the invention: SYSTEM OF PROCESSING FOR ANALYSIS AND OPTIMIZING TRANSPORT ROUTES

[0001] Technical field

[0002] The invention relates to a processing system dedicated to the analysis of transport operations.

[0003] State of the art

[0004] In the field of advanced logistics, goods are frequently transported between various countries. The points of origin and destination are generally very specific, such as a collection site belonging to company A and a delivery site located at company B. This creates a complex and extensive transport network.

[0005] Furthermore, the nature of current shipments relies primarily on time-sensitive logistics, combining pre-negotiated contracts with carriers (contracted carrier) and ad-hoc purchasing solutions on the spot market to meet on-the-fly demands. This necessitates an optimized offering for each individual situation.

[0006] The requirements for time sensitivity translate into increased difficulties in the design of transport plans. Indeed, these plans must be developed in real time, making the forecasting of future transport needs particularly complex. This unpredictability limits the ability of current tools to anticipate and optimize plans effectively to meet operational and economic constraints.

[0007] The routes taken are likely to change frequently due to various factors: specific customer requirements, traffic constraints (such as traffic jams or roadworks), availability of means of transport, etc. This evolving and complex nature considerably complicates the effective planning of advanced logistics services.

[0008] To ensure efficient management and prepare for the installation of tachometers in trucks over 2.5 tonnes, a logistics provider must optimize the use of co-loading, which consists of grouping different shipments onto the same means of transport. However, this approach is only viable if certain portions of the routes of the shipments concerned coincide or can be adjusted to coincide. The increasing complexity and extent of transport networks make this identification of common segments increasingly difficult.

[0009] Furthermore, existing technologies are primarily based on a classic milkrun-type approach to optimizing repetitive routes or on systems based on central points such as logistics hubs. While these methods are suitable for stable and predictable transport scenarios, they show their limitations when faced with dynamic logistics flows or the need for immediate adaptation to transport disruptions. These limitations prevent fully optimized management of time-sensitive logistics flows or combined transport solutions.

[0010] In the context of co-loading, as well as for other applications, it is essential to effectively cross-reference and integrate data related to different routes in order to identify relevant relationships. This process relies on the rigorous analysis and processing of a large volume of routes.

[0011] Technical problem

[0012] The technical problem arises from the efficient processing of journey information to find co-loading probabilities.

[0013] Current transport planning solutions rely on classical optimization approaches, often based on vector models or regression analyses. However, these approaches show their limitations in complex and dynamic scenarios. Indeed, when simplified to the extreme, they lack precision and fail to identify crucial opportunities, such as co-loading.

[0014] Consider the example of a route with a collection point in Valencia and a delivery point in Rome: according to a classical vector analysis, this route would not be identified as compatible with another route between Bordeaux and Venice. However, by examining the actual routes taken, it becomes clear that with a few minor detours, a shared load could be feasible. Multiconsecutive vector approaches could improve the analysis by incorporating complex route combinations. However, these approaches significantly increase data processing requirements, making their application economically and technically unviable on a large scale.

[0015] Added to this is the challenge posed by time constraints. The nature of current shipments, often sensitive to the time factor, requires taking into account an additional dimension: that of time. Carriers must therefore integrate this temporal variable to meet the requirements of rapid delivery. However, this added complexity makes computer processing even more expensive and sometimes impossible to perform within the timeframes required by operational needs, particularly in real time.

[0016] In summary, classical approaches are not suitable for responding to dynamic logistical flows, integrating both spatial and temporal constraints and Complex opportunities such as co-loading justify the need for innovative solutions capable of overcoming these limitations while offering processing times compatible with immediate logistical requirements.

[0017] The invention relates to a processing system designed to analyze transportation data. The term "analysis" encompasses various levels of interpretation, ranging from simple information extraction to complex processing to generate usable data from raw data. This system, whose complexity can vary, typically includes a processing unit and a storage device, often based on a standard computer configuration such as a personal computer (PC). Although the focus is on the transport of goods, the system is also applicable to the transport of people or animals.

[0018] The described system offers an innovative alternative that begins with an initial simplification: all transport data is considered two-dimensional (ignoring altitude variations in road transport). Each shipment is represented in its finest detail by a sequence of points on a two-dimensional plane. This sequence is then transformed (or discretized) into a one-dimensional sequence of nodes via a neural mapping process.

[0019] The nodes are defined from a limited set of possibilities, pre-established through the analysis of historical data. Thus, each transport is described as a linear sequence of passage points (one-dimensional dimension) to which is associated a second discrete dimension: a passage time window.

[0020] This approach significantly reduces the complexity of the data to be processed while allowing the integration of temporal and spatial constraints into the analysis. It constitutes a solution adapted to the needs of dynamic expeditions, meeting the requirements of real-time optimization.

[0021] The system integrates data relating to a set of "nodes" located on the map, each node being associated with a proximity zone. This data, often expressed as coordinates, corresponds to strategic points such as warehouses, train stations, airports, or major intersections. These nodes are defined before the analysis and are independent of a specific sequence, although they may be identified by a name or a number. The proximity zone of a node is a region around it, generally circular and of uniform size, but it may have specific shapes or dimensions to take into account geographical constraints or the relative importance of the node.

[0022] To process the data, the system has an interface to receive this information and memory to store it. It is designed to perform a mapping procedure for each path, generating a sequence of nodes crossed by the path in their respective proximity zones. By traversing a path, the The system identifies nearby nodes and includes them in a sequence based on defined criteria. These sequences, or derived information, are then stored for future use.

[0023] According to the invention, the processing system is designed to perform a mapping procedure for each path. This procedure consists of generating a sequence of nodes corresponding to the proximity zones traversed by the path. Specifically, the system "traverses" the path by following its orientation and evaluates at each step whether it enters the proximity zone of a node. If not, the node is ignored. Conversely, if the path enters the proximity zone of one or more nodes (the zones may overlap), it is then considered "near" these nodes. The system compiles a sequence of nearby nodes, although this sequence can be refined according to additional criteria. This sequence is generally saved for later use, or information extracted from it is stored instead.

[0024] Mapping provides a synthetic view of the route in relation to nodes, which are often strategic points on the map. It is not limited to a crude simplification reducing the route to a small number of reference points. When a route is located in the vicinity of a node, the node is "associated" with it, even if the route does not pass directly through it. This can, for example, allow for route optimization so that future transport passes through this node, thus facilitating potential sharing with other nearby routes. This is one of the many possible applications of the results obtained from the mapping procedure.

[0025] The mapping procedure associates a path (described in two or more dimensions) with a one-dimensional sequence of nodes. Although the coordinates of the nodes are expressed in the same dimensional space as the path, the sequence itself can be interpreted as a one-dimensional object. This process of dimensionality reduction or data compression optimizes resources in terms of computation and processing time.

[0026] Each route is generally defined by a sequence of waypoints, spaced at equal distances. A distance between 5 and 20 km is often useful, with a preference for 10 km. Each waypoint is represented by coordinates, usually latitude and longitude. During the mapping procedure, each waypoint is analyzed to determine whether it lies within the proximity zone of one or more nodes. By reducing the number of waypoints, the memory required to store the routes is minimized, and the mapping execution is accelerated.

[0027] In the event of overlapping proximity zones, the mapping procedure could produce ambiguous or complex results, for example if a path crosses multiple zones can be accessed simultaneously. To avoid this, the system prioritizes the nearest node and only adds it to the sequence. This selection is based on the minimum distance between the waypoint and the nodes under consideration. Thus, more distant nodes, although nearby, are excluded because their logistical importance is lower.

[0028] The mapping can theoretically take into account all the nodes on the map, but pre-selecting the relevant nodes reduces processing time. For example, for a route from southern France to northern Spain, it would be unnecessary to include nodes located in Germany. The system is therefore configured to select, before mapping, a subset of regional nodes based on the geographical area delimited by the starting and ending points. This subset optimizes the analysis without including superfluous nodes. If all the nodes are selected, the subset becomes identical to the initial set.

[0029] Even when limiting the selection to the nearest node, a complex or winding path could produce an irregular node sequence, such as a "zigzag". To avoid this, the system is configured to exclude a node from the subset of regional nodes once a waypoint has been identified within its proximity zone. This node will nevertheless be added to the final sequence if the criteria are met. This ensures that each node is considered only once per path.

[0030] Various methods can be used to define the relevant regional nodes. The area to be considered must be continuous and include the starting and ending points of the route. A simple and efficient method is to select the nodes located within an ellipse defined by the starting and ending points as foci. The ellipse can be nearly circular, but its width must remain significant, for example, at least 20% of the distance between the two points, in order to avoid excluding too many nodes and obtaining insufficient results.

[0031] Another method consists of using a circle centered at the midpoint of the segment connecting the starting point and the ending point. The diameter of this circle is greater than the distance between these two points in order to include more potentially relevant nodes.

[0032] Travel data is generally generated by navigation software, which calculates an optimal route between a starting point and a destination point provided by a user. This software can operate directly on the processing system or on an external system, which then transmits the information to the system. In addition, position sensors, such as GPS devices, can provide data on the route actually followed. This data makes it possible to check for deviations from the planned route. GPS sensors, often integrated into means of transport or mobile devices such as smartphones, can transmit this data in real time to the processing system via mobile technologies or record it for later transfer. However, the means of transport could It can also be equipped with specific components, such as sensors and a transmitter dedicated solely to transmitting position data to the processing system. Another approach would involve a user, such as a driver, defining the starting point and destination directly from the vehicle. This information would then be transmitted to the processing system, which would calculate the route.

[0033] One of the main advantages of the processing system lies in its ability to analyze the importance of connections between specific pairs of nodes, that is, to determine the magnitude of transport flows between their proximity zones. For each pair of nodes, the system is designed to compile the route information present in all sequences of nodes containing that pair. For example, for a pair "A" and "B", the system identifies all sequences where "B" follows "A" (or vice versa, depending on the relevant order) and sums the corresponding information.

[0034] In one variant, only routes where the loading point belongs to node A and the delivery point to node B are considered. One possible analysis involves counting the frequency of occurrence of a given pair in the sequences. In this case, the system acts as a counter, increasing by one each time a specific node pair is detected. This counter provides a quantitative representation of the flows between the two nodes.

[0035] In addition or as an alternative, the system can also integrate transport-specific parameters, such as the volume, weight, or value of the goods transported. This data allows for a better assessment of the relative importance of a flow. For example, it is crucial to know not only the number of trips between "A" and "B," but also whether they involve small or large quantities of goods. This information enriches the logistics analysis by providing a more nuanced understanding of the exchanges between pairs of nodes.

[0036] The processing system also makes it possible to assess the number of transport resources needed for a particular node. This is done by counting how many times a node (and its surrounding area) is used as a starting point for pickups. This monitoring can highlight seasonal or weekly fluctuations in transport requirements. For example, by recording this data for each week of the year, the system can predict requirements for a given period. By comparing the forecasts with actual data, it is possible to assess the predictability of flows and identify long-term trends, such as variations between two consecutive periods (weeks or years). This allows for more precise planning of logistics resources.

[0037] A significant advantage of this system lies in its ability to identify co-loading or reloading opportunities. These opportunities arise when a route passes near a location scheduled for loading, unloading, or reloading. For example, if a first transport is near a location where a second transport needs to load cargo, the system can suggest modifying the first transport's route to carry out this pickup, thus optimizing resources. In this case, the two journeys can be combined and completed by a single means of transport.

[0038] To make this possible, the system identifies the nodes in a sequence associated with loading operations. It then plans an alternative route to include these nodes and perform the necessary logistical operations. However, the system or the user may reject this option if it proves inefficient or incompatible with other criteria.

[0039] The system also optimizes the use of transshipment points, which allow a journey to be divided into shorter segments, completed by different modes of transport. This strategy is often more cost-effective than making a long journey with a single vehicle. When a sequence of nodes contains a transshipment point, the system can plan alternative routes including these nodes to carry out transshipment operations. These divisions make it possible to coordinate multiple loading or unloading operations while minimizing the distances traveled by each vehicle.

[0040] Several factors are taken into account when deciding on the use of one or more transshipment points: • Distances: The segments between transshipment points must be neither too long nor too short. • Regulatory compliance: Constraints such as the maximum number of driving hours are respected. • Time efficiency: An excessive number of transshipment points or prolonged delays can compromise the quality of service. • Coordination of co-loading: When several vehicles need to meet at a point to reload or transfer goods, timing is crucial.

[0041] Finally, even if these optimized routes are planned, they can be rejected by the system or a user if they do not meet the profitability or efficiency criteria.

[0042] To identify the most frequently used transshipment points, the system calculates a recurrence factor for each. This factor corresponds to the number of times a transshipment point has been used over a given period. This calculation takes into account It only counts journeys involving a transshipment operation and excludes journeys that merely pass through the vicinity of such a point. Points with a high recurrence factor can be prioritized for new, optimized journeys, as mentioned previously.

[0043] In most cases, the use of transshipment points involves reloading operations carried out by means of transport belonging to the same logistics provider. However, further optimization can be achieved by integrating third-party transport, i.e., services provided by other logistics providers. In this case, the schedules and routes of the third parties are predetermined, but their use can significantly reduce the resources required.

[0044] Thus, the system can be configured to plan a route including third-party transport if a node in the sequence corresponds to a departure or arrival point for that transport. However, the system can discard this option in favor of the initial route if the latter proves more advantageous. Normally, the third-party option is selected if the total combined route (pre-shipment + third-party transport + post-shipment) is shorter than the original direct route, even when simplifying the comparison to geographical distances. Furthermore, this option is abandoned if the schedules do not match.

[0045] When two transports take place in a similar area and follow compatible directions, it is relevant to check whether they can be combined into a single transport. The corresponding means of transport could adjust its route to incorporate the second journey.

[0046] In this context, the system is designed to analyze the starting and ending points of the journeys (as nodes). If a node of one journey corresponds to a starting or ending point of another journey, the system proposes a new combined route. This new route encompasses the starting and ending points of both modes of transport, allowing a single means of transport to complete them simultaneously.

[0047] The described processing system can be modified in several ways. For example, the number of nodes can be reduced to one or two per country, and in this case, the proximity zones of these nodes must be sufficiently large to allow for a large-scale analysis of flows between countries. This makes it possible to calculate the total flow from one node to another and vice versa. An analysis of inbound and outbound flows can reveal the importance of a node, as well as the balance between the two flows. It is also possible to compare the frequency of use of a node as a collection point versus its use as a delivery point, which can help anticipate transport needs at that location. A similar analysis can be carried out for several nodes in a given region, for example, a country, in order to obtain a "regional balance". These results can also be analyzed separately for each service provider. Logistics services. The results of this analysis can be visualized as maps where arrows of varying sizes indicate flows, and unbalanced nodes or regions are colored differently. Finally, flow analysis also helps identify optimal locations for transshipment points. The zones of proximity to nodes can have various shapes or radii, depending on the needs of the analysis.

[0048] The invention also relates to a processing system designed to analyze transportation. This system is configured to receive data defining several routes on a two-dimensional map. Each route represents a piece of transportation. The system can also automatically determine the route between a given starting point and a given ending point, autonomously determining the itinerary.

[0049] Features of the invention - The processing system for transport analysis is configured to: • Be equipped with data representing multiple routes on a map in minus two dimensions, each journey representing a transport, • Have data defining a set of nodes on the map, each node having a proximity zone, • Execute a mapping procedure for each path, generating a sequence of nodes crossed by the proximity zones of the path; each path is defined by a sequence of waypoints, and the mapping procedure includes checking each waypoint to determine if it is in the proximity zone of at least one node. - If a waypoint is located in several proximity zones, the processing system only adds the closest node to the sequence. - The processing system is configured to select, before each mapping procedure, a subset of regional nodes from the set of nodes, and to only take into account the proximity areas of the regional nodes. - If a waypoint is located within the proximity zone of a node, that node is removed from the subset of regional nodes. - The processing system selects a node as a regional node if it is located inside an ellipse defined by an initial point and an end point of the path. - For each pair of nodes, the processing system adds information about the paths of all node sequences containing that pair of nodes - The processing system is configured to count the number of occurrences of each pair of nodes. - The processing system is provided with a characteristic transport parameter and adds this parameter to the journey information. - If a sequence of nodes in a path includes a node intended for a loading operation, the processing system organizes another path that includes this node. - If the sequence of nodes includes a node corresponding to a transshipment, the processing system organizes another path which includes this node and corresponds to a transshipment operation at the level of this node. - If the sequence of nodes includes a node corresponding to an initial or final point of a third-party transport, the processing system organizes another route that includes this third-party transport. - The system is provided with starting and ending points of paths as nodes, and is configured so that if the sequence of nodes of a first path includes a node corresponding to a starting or ending point of a second path, the processing system organizes another path including the starting and ending points of both paths. Brief description of the drawings

[0050] Examples of embodiments of the invention are presented with accompanying drawings, including: • [Fig.1]: Schematic view of a means of transport and a processing system according to the invention. • [Fig. 2]: Diagram illustrating a system mapping procedure treatment, at a first stage. • [Fig.3]: Diagram illustrating the second-step mapping procedure. • [Fig. 4a]: Diagram illustrating a transport route and several points of transshipment. • [Fig. 4b]: Diagram illustrating another alternative route for the journey of transport of the [Fig.4a]. • [Fig.5a]: Diagram illustrating a transport route and a third transport. • [Fig. 5b]: Diagram illustrating another alternative route for the journey of transport of the [Fig.5a]. • [Fig.6a]: Diagram illustrating two transport routes. • [Fig. 6b]: Diagram illustrating another alternative route for the journeys of transport of the [Fig.6a].

[0051] Description of preferred embodiments

[0052] Figure 1 shows a processing system 40 according to the invention. This system comprises a processing unit 41 (for example, a central processing unit of The processing unit (PMU) of a personal computer is connected to a memory device 42 (such as main memory or the computer's hard drive). The processing unit can read and store data in this memory device. Furthermore, the processing unit 41 is connected to a first interface 43 and a second interface 44 to receive external data. The first interface 43 can be an input device such as a keyboard or mouse, or a connection to a data source, for example, via a network. The second interface 44 is configured to receive signals, whether wired or wireless.

[0053] The processing system 40 is designed to receive data relating to several routes 1 on a map. Each route describes a transport operation carried out by a means of transport 100, schematically represented as a truck in [Fig. 1]. The data can be entered via the first interface 43, but the system can also generate certain route details itself. For example, a starting point 2 and an ending point 8 are specified by the user via the interface, and the route details are calculated by the processing system using known navigation algorithms. These algorithms can also be executed outside the system.

[0054] The truck 100 is equipped with a GPS positioning device 101 to determine its position. The position data obtained is transmitted to the second interface 44 via a wireless transmitter 102. This signal is generally relayed by terrestrial link or satellite. Although the GPS and transmitter are shown as components of the truck, they could also be part of a smartphone with GPS functionality. Communication with the transmitter and the second interface is used to verify the truck's position and, if necessary, to adjust its route.

[0055] The data sent by the transmitter consists of a series of coordinate pairs corresponding to the truck's route. To simplify the analysis, the truck's position is stored only at 10 km intervals, meaning that only these 10 km segments are considered. The system can also determine these waypoints by interpolating the data.

[0056] Thus, route 1 is represented on a 2D map by a series of waypoints 2-8, going from the starting point 2 to the arrival point 8, where the truck performs the loading and unloading operations. In addition to these points, a set of nodes 10-22 is defined, each node representing a significant point such as a city, a transshipment point, a warehouse, a train station, or an airport. For each node, a circular proximity zone 30 is defined, with a radius of 15 km.

[0057] At the beginning of the analysis, the system determines the regional nodes 10-19 that are relevant for the path, selecting them within an ellipse 31 whose The centers are the starting point 2 and the ending point 8. Nodes located outside this ellipse, such as nodes 20-22, are not taken into account for this route.

[0058] The processing system checks each waypoint to determine if it is within the proximity zone of a regional node. For example, the path between the starting point 2 and the first waypoint 3 does not cross the node's proximity zone, but waypoint 4 is within the proximity zone of node 15. This node is therefore selected to begin a sequence of nodes and is then removed from the list of regional nodes to avoid being considered again.

[0059] As the path progresses, each new waypoint is analyzed to see if it falls within the proximity zone of another node. After several steps, the system adds new nodes to the sequence and removes them from the set of regional nodes. This continues until the path reaches its endpoint.

[0060] Finally, the sequence of nodes selected for the route is completed, representing the path followed by truck 100 and the significant points encountered along the way. This process connects each waypoint to a node and completes the mapping of the route on the map.

[0061] The processing system 40 stores the sequence of nodes in the memory device 42. This mapping procedure is applied to several paths 1 corresponding to different routes of means of transport 100. Subsequently, the system generates a "flow matrix," where each row and each column corresponds to a specific node. Initially, all values ​​in the matrix are equal to zero. The system then checks each stored sequence of nodes. If, for example, a sequence of nodes contains the consecutive nodes 11 and 17, the value at the intersection of row 11 and column 17 of the matrix is ​​incremented by one. It is also possible to create this matrix as the node sequences are identified, without necessarily storing them beforehand.

[0062] The generated flow matrix allows for the analysis of important characteristics such as the flow between nodes, the overall flow from one node to other nodes, the balance of flows entering and leaving a node, and so on. Characteristic values, such as the weight of the transported goods, can also be included. In this case, instead of simple integer values, the matrix will contain values ​​corresponding to the total weight of the goods. For example, if a sequence of nodes includes certain other nodes, the value in the matrix is ​​increased by the weight of the goods involved in the transport.

[0063] Figure 4a illustrates a transport route 1, where only the starting point 1 and the ending point 8 are shown, while the intermediate points are omitted. Three transshipment nodes 23-25 ​​are also shown. The route passes through the proximity zones 30 of two nodes, 23 and 24, which are therefore included in the node sequence of the route. 1, according to the mapping procedure described previously (possibly including other nodes not shown here). The system then analyzes the possibility of using these nodes as reloading points. For example, another route (1b) is planned to include transshipment points 23 and 24. This route is usually operated by three separate modes of transport. Furthermore, the processing system can plan other routes that include only one of the transshipment points, depending on the possibility of co-loading or other factors.

[0064] Figures 5a and 5b illustrate the use of the system to organize transport by third parties. A transport route 1, from the starting point 2 to the final point 8, is shown, with a node 26 representing the starting (pick-up) point and a node 27 representing the delivery point of transport carried out by a third-party provider. This third-party transport is represented by a dashed line. The system detects an opportunity to use this third-party transport if route 1 enters the proximity zone of node 26. Consequently, another route 1b is planned, modifying the itinerary to include nodes 27 and 28. The itinerary passes through node 26 to reload the goods into the third-party transport before continuing to the final point 8.The choice of the optimal route may depend on several criteria, such as the total distance travelled or the associated costs, and may be made by the processing system or by a user after a pre-selection.

[0065] Figures 6a and 6b illustrate an example of co-loading without intermediate changes of means of transport. In [Fig. 6a], a first transport route 1, going from the starting point 2 to the ending point 8, and a second route 1a, going from the starting point 2a to the ending point 8a, are shown. In this case, the starting and ending points 2a and 8a of the second route are treated as nodes in the mapping of the first route 1. Similarly, the points of the first route 1 are integrated into the mapping of the second route 1a. The fact that the first route 1 passes through the proximity zone 30 of the starting point 2a of the second route 1a indicates that the two routes can potentially be combined. Therefore, a new route 1b is planned, as shown in [Fig. 6b]. The route of this route 1b is modified to include the starting points 2 and 2a, as well as the ending points 8 and 8a. This route goes from the initial point 2 to the final point 8, via 2a and 8a.The system also plans other routes (not shown here), in which the sequences of points 2, 2a, 8, and 8a may be different. It is clear that the total distance of route 1b is less than the sum of the individual routes 1 and 1a, and other sequences may also lead to a reduction in distances compared to the separate routes. However, the sequence shown in [Fig. 6a] is the most optimal.

[0066] It is important to note that it is possible to combine different embodiments of the processing system of the invention, as shown in Figures 4a to 6b. The The system can search for options to use transshipment, third-party transport, and a combination of transport simultaneously.

Claims

Demands

1. A processing system (40) for transport analysis, the processing system (40) being configured to: • Have data representing several paths (1, la) on a map in at least two dimensions, each path representing a transport, • Have data defining a set of nodes (10-27) on the map, each node having a proximity zone (30), • Execute a mapping procedure for each path (1, la), generating a sequence of nodes (11, 15, 17, 19) crossed by the proximity zones (30) of the path, characterized in that, for a pair of nodes (10, 27), the processing system is configured to add information concerning the paths (1, la) of all node sequences that contain said pair of nodes (10, 27), in that the processing system is configured to count the number of times said pair of nodes (10, 27) appears,This allows for the generation of a flow matrix in which each row and column corresponds to a node, and the processing system is configured to analyze the resulting flow matrix based on characteristics such as the flow from one node to another, the overall flow from / to a node to / from all others, the flow balance to and from a node, and thus enable the organization or planning of optimized transport operations.

2. System according to claim 1, characterized in that each path (1, la) is defined by a sequence of waypoints (2-8, 2a, 8a), and that the mapping procedure includes checking each waypoint (2-8, 2a, 8a) to determine if it is located in the proximity zone (30) of at least one node (10-27).

3. System according to claim 2, characterized in that, if a waypoint (2-8, 2a, 8a) is located in several proximity zones (30), the processing system (40) adds only the nearest node (10-27) to the sequence.

4. System according to claim 2 or 3, characterized in that the processing system (40) is configured to select, before each mapping procedure, a subset of regional nodes (10-19) from the set of nodes (10-27), and to take into account only the proximity zones (30) of the regional nodes (10-19).

5. System according to claim 4, characterized in that, if a waypoint (2-8) is located within the proximity zone (30) of a node (10-27), that node (10-27) is removed from the subset of regional nodes (10-19).

6. System according to claim 4 or 5, characterized in that the processing system (40) selects a node (10-27) as regional node (10-19) if it is located inside an ellipse (31) defined by an initial point (2, 2a) and an end point (8, 8a) of the path (1, la).

7. System according to claim 1, characterized in that the processing system (40) is provided with a characteristic transport parameter and adds this parameter to the journey information.

8. System according to any one of claims 1 to 7, characterized in that, if a sequence of nodes of a path (1, la) includes a node (2a, 23, 24, 26) intended for a loading operation, the processing system (40) organizes another path (la) which includes this node (2a, 23, 24, 26).

9. System according to claim 8, characterized in that, if the sequence of nodes includes a node (23, 24) corresponding to a transshipment, the processing system (40) organizes another path (1b) which includes this node (23, 24) and corresponds to a transshipment operation at the level of this node (23, 24).

10. System according to any one of claims 10 to 11, characterized in that, if the sequence of nodes includes a node (26) corresponding to an initial or final point of a third-party transport, the processing system (40) organizes another path (1b) which includes this third-party transport.

11. System according to any one of claims 8 to 10, characterized in that the system is provided with initial points (2, 2a) and final points (8, 8a) of paths (1, la) as nodes, and is configured such that if the sequence of nodes of a first path (1) includes a node (2a) corresponding to an initial or final point of a second path (la), the processing system (40) organizes another path (1b) including the initial points (2, 2a) and the final points (8, 8a) of the two paths (1) and (la).

Images