Philippine
Road Networks

The UPLB Road Network

The Road Network of the academic campus of the University of the Philippines Los Baños (UPLB) is the initial road network dataset available in this repository. The map image presented in Figure 1 shows UPLB's roads on the academic campus as seen (overlaid with blue lines for visual identification) on the OpenStreetMap website.

In creating this dataset, we focused exclusively on the roads within the main academic campus and deliberately excluded the roads in specific areas such as the UPCO Faculty Housing, the DTRI campus, and the research campuses of IPB and BIOTECH. This targeted approach ensures that the dataset accurately represents the core academic infrastructure of UPLB without including peripheral areas that might not be relevant to most users, especially the students.

We extracted the street data from OpenStreetMap, which serves as the foundation for our road network. OpenStreetMap is a valuable resource for obtaining up-to-date and detailed geographic information, and its open-source nature allows for comprehensive and accurate mapping. By utilizing this data, we ensure that the road network representation is current and reflects the actual layout of the campus roads.

This initial dataset is a critical resource for various applications, including campus planning, navigation, and infrastructure development. By providing a focused and accurate depiction of UPLB's academic campus roads, we enable users to perform detailed analyses and make informed decisions based on reliable data.

Fig 1. The UPLB campus map as seen in OpenStreetMap.

Graph Extraction

The geographical data of the blue lines in Figure 1 was extracted to transform the image into the map image depicted in Figure 2 where it shows UPLB's road network in white solid lines. The geographic boundary of the academic campus is shown in light green line. The roads shown in Figure 2 are the set union of the walkable, drivable, and bikable road networks. Walkable roads are clear, sometimes paved, pathways the pedestrian traffic would normally use. This includes both public, private and special access walkways (e.g., PWD ramps). Drivable roads are roadways where vehicular traffic is allowed, while bikable roads are pathways reachable by bicycles and micro-mobility transportation devices (e.g., electric scooters and skateboards).

Figure 3 shows the same map image in Figure 2 with the node data in yellow overlaid on each intersections, dead-ends, and along road segments. The nodes along road segments capture the curvature (i.e., geometry) of the segment but is not part of the node set of the road network. The extra node data for road segment geometry information need to be removed from the network. This was done by selecting all nodes whose degree d is exactly two (i.e., d = 2) and removing them from the node set. The two edges that were adjacent to the removed node were joined as one edge. Thus, for any k-node road segment in this figure, there are k - 2 internal nodes of degree d = 2, and k - 1 edges. After removing the k - 2 internal nodes and k - 2 edges, what remains are two nodes and one edge.

The order of the road network (i.e., the number of nodes n or the cardinality of the set V) depicted in Figure 3 was found to be n = 2,497, while the size of the road network (i.e., the number of edges m or the cardinality of the set E) was found to be m = 6,592.

Fig. 2. The network of UPLB streets.

Fig 3. The UPLB road network overlaid with node data (n=2,497, m=6,592).

Figure 4 below shows the road network with the node data that define the road segment geometry removed. Even if we removed these nodes from the node set V, we kept these node data as respective geometric attributes of the edges so that we can accurately plot the road segments the edges represent. Lines in red represent the edges of a small disconnected component of the road network. A disconnected component has nodes that can not be reached from most nodes in the network. Disconnected components must also be removed from the network. The road network in this figure has n = 1,077 and m = 2,978.

The road map depicted in Figure 5 shows the largest connected component (LCC) of the UPLB road network. Here, n = 1,063 and m = 2,964. The graph information of this road network is the one made available in this web resource.

Fig 4. The UPLB road network with the nodes that define the road segment geometry removed (n=1,077, m=2,978). Road segments in red are edges in smaller connected component.

Fig 5. The largest connected component of the UPLB road network (n=1,063, m=2,964).

Walkable Road Network

The walkable network was extracted from the LCC of UPLB's road network. Figure 6 shows the map image of the network with the node data in blue, the edge data in black, and the geographic bounds in green. Notice that some edges outside of the bounds were included. These edges are walkable private pathways that cater to UPLB students and constituents who live in private residences along the road segments the edges represent. The network has n = 1,014 and m = 2,936. Because pedestrians can physically (and legally) go either way on any road segment, the road network is undirected. In undirected road network, we treat the edge (i, j) to be the same as the edge (j, i), ∀ i,j ∈ V. Symbolically, (i, j) = (j, i).

Fig 6. The UPLB walkable network (n = 1,014, m = 2,936).

Fig 7. The isochrone map of UPLB walkable network with respect to the UPLB Gate. Nodes with the same color are reachable from the UPLB Gate via the shortest route within a certain walking time for a 4.5km/hr average walking speed.

Travel Time Analysis: UPLB Gate

The UPLB walkable network is particularly useful for providing pedestrian mobility information. For instance, if a person needs to walk from the UPLB Gate (longitude of 121.24343 and latitude of 14.16761) to any location within the academic campus via the shortest route possible, one might wonder how long it will take to reach the destination if the average walking speed is 4.5 km/hr. The isochrone map in Figure 7 illustrates the UPLB walkable network with colored nodes, which provides the answer to this question in a visual manner.

In Figure 7, maroon nodes represent locations reachable from the UPLB Gate within a 5-minute walk. Yellow-orange nodes indicate destinations reachable within 10 minutes, light green nodes within 15 minutes, light blue nodes within 20 minutes, and purple nodes within 25 minutes. These color-coded nodes visually indicate the walking times from the UPLB Gate to various points across the campus, offering a clear and practical guide for pedestrians.

Since the walkable network is bidirectional, the same map in Figure 7 can also be used to determine how long one would have to walk from any starting point within the academic campus to the UPLB Gate. This bidirectional feature ensures that the network provides comprehensive mobility information, regardless of the direction of travel.

Travel Time Analysis: Copeland Gym

The Edwin B. Copeland Gym is one of the most visited facilities on UPLB's academic campus. This high level of visitation is primarily because almost all undergraduate students are required by their academic programs to take 12 units of Physical Education subjects. To provide these students with relevant information for visiting the gym, we can recreate the isochrone information shown in Figure 7, but this time with respect to the location of the gym, which is at a longitude of 121.24274 and a latitude of 14.15668.

This recreation is shown in Figure 8. By doing this, we offer a detailed visual representation of walking times to the gym from various points on campus. Similar to Figure 7, the nodes are color-coded to indicate the time it takes to walk to the gym. Maroon nodes represent locations reachable within a 5-minute walk, yellow-orange nodes within 10 minutes, light green nodes within 15 minutes, light blue nodes within 20 minutes, and purple nodes within 25 minutes.

This information is invaluable for students who need to plan their routes to the gym efficiently. By understanding the walking times from different locations on campus, students can better manage their schedules and ensure timely arrivals for their Physical Education classes. The isochrone map provides a practical and user-friendly guide to navigating the campus, highlighting the accessibility of the gym from various starting points.

Fig 8. The isochrone map of UPLB walkable network with respect to the Copeland Gym.

The UPLB walkable network serves as a valuable tool for students, faculty, and visitors, helping them plan their walking routes efficiently. By understanding the walking times between different locations on campus, individuals can make informed decisions about their routes, manage their time better, and ensure timely arrivals at their destinations.

Drivable Road Network

Fig 9. The UPLB drivable network (n = 77 and m = 185).

Extracted from the LCC is the drivable network whose map image is shown in Figure 9. Road segments that lead to parking spaces were removed from the network, such as those in the Animal Science-Veterinary Medicine subcampus and Development Communication-CAFS Administration sub-campus. Likewise removed are access road ways to UPLB's experimental farms where only tractors, heavy machines, and UPLB vehicles used for the conduct of field experiments are allowed. This network has n = 77 and m = 185.

UPLB's vehicular traffic along the drivable roads are implemented in either a one-way or a two-way scheme and these schemes must be reflected in the network: Some edges will represent for a one direction traffic while some for two. Thus, this network is a directed network where edge (i,j) is different from edge (j,i), ∀ i,j ∈ V.

Even if we exclude the edges that represent road segments to and in the experimental fields of UPLB's research campuses, we can extract the respective road networks of such campuses for whatever academic, research or management purposes the network would serve.

Centrality Analysis

The travel time analysis for the UPLB walkable network shown in Figures 7 and 8 can also be conducted for the corresponding drivable network. To perform this analysis, one simply needs to adjust the average travel speed to the speed limit, depending on the vehicle used. This allows for a more accurate representation of travel times for drivers within the campus.

One network metric that is particularly useful for route planning, whether for personal, special event, or institutional purposes, is nodal betweenness centrality. The betweenness centrality of any node i is the ratio of the number of shortest paths that pass through i to the number of all shortest paths in the network. A high nodal betweenness centrality indicates that a node is part of many shortest paths, suggesting it is a commonly traversed route. This often means the node is a busy intersection or a crucial point in the network. Conversely, the betweenness centrality of dead ends is close to zero, as they do not lie on many paths between other nodes. Note that some nodes may be wrongly considered as dead-ends but actually are parts of road segments that were cut by the geographic boundary lines.

Figure 10 illustrates the UPLB drivable network, with colored nodes representing their respective betweenness centralities. This visualization helps identify key intersections and pathways that are critical for efficient route planning. High betweenness centrality nodes are highlighted, showing the most traversed points within the network. This information is invaluable for optimizing traffic flow, managing congestion, and planning routes for various purposes.

By understanding the betweenness centrality of nodes in the drivable network, planners can make informed decisions to enhance transportation efficiency and improve the overall mobility within the campus. This analysis supports a wide range of applications, from daily commuting to organizing large-scale events, ensuring that routes are planned effectively to accommodate traffic demands.

Fig 10. the UPLB drivable network with colored nodes representing the respective nodal betweenness centralities of the nodes.

Bikable Road Network

Fig 11. The UPLB bikable network (n = 369, m = 870).

Similar to the extraction of the walkable and drivable networks, we also extracted the bikable network from the LCC, as shown in Figure 11. This network is undirected reflecting the bidirectional nature of bicycle and micro-mobility transportation paths. In this bikable network, n = 369 and m = 870.

By extracting the bikable network, we provide a detailed representation of the bicycle pathways within the UPLB campus. This network includes paths that are accessible and safe for cyclists, promoting sustainable and healthy transportation options. The undirected nature of the network accurately reflects the real-world conditions where cyclists can travel in both directions on the same path.

Understanding the bikable network's structure is crucial for various analyses. These metrics help in assessing the connectivity and accessibility of bicycle paths, identifying potential gaps or areas for improvement, and supporting campus planning and infrastructure development with the inclusion of cycling, among other considerations.

Page updated

Report abuse

PhilippineRoad Networks