By ELMO DOMINO R. JOSE
A geospatially well-constructed community includes accessible facilities. For instance, a previous study found fast-food chains to be situated closely to each other, yet at the same time, very far from each other in far-flung areas. The presence of universities and hotels increase the likelihood of finding such a facility. Other examples of these commonly accessed facilities are vaccination sites, gas stations, hospitals, and automated teller machines (ATM’s).
Though the prevalence of digital banking and financial technology services, such as GCash, have made money transactions convenient, some sectors remain reliant on physical banking like ATM’s and over-the-counter transactions. For instance, workers and students who receive their compensation or stipends through banks, like the Land Bank of the Philippines—the largest government-owned bank, must travel first to withdraw cash.
Depending on their residence, banking customers are faced with the task of commuting to reach these ATM facilities. Hence, we must ensure that these facilities are accessible to them and perhaps even easily walkable. Given that, we are asked the question: How can we gauge accessibility?
In our study, we aimed to gauge accessibility using two key factors: determining the coverage area and the nearest alternative of a Land Bank ATM facility. In this case, we analyzed the locations of Land Bank ATM’s in the National Capital Region.
Knowing the coverage area gives us a glimpse of how many users are being serviced. Moreover, determining the nearest backup facility is crucial to determine the overall resilience of the Land Bank ATM network as these machines are prone to going offline, and servicing them takes a while to be accomplished.
To determine the coverage area of each ATM, we demarcated the areas using a mathematical approach called Dirichlet tessellations. Essentially, Dirichlet tessellations partition an entire area into smaller regions. This partitioning is done such that every region encloses a single ATM facility. For every point in the enclosed region, its distance is closest to its own corresponding ATM facility than any other branch. Given this presupposition, the coverage areas of ATM’s are demarcated.
Next, we determined the nearest backup of every facility to determine how walkable and robust the entire system is. We determined this using proximity analysis, which identifies the nearest alternative branch, in the event that the usual ATM site is offline for a user.
Smaller tessellation area means that a Land Bank ATM is more accessible. Larger areas mean that people must travel farther to access such facilities. Our analysis showed that 68 percent of the tessellations have a coverage area of around 1.25 km2 or less, which can be easily accessed by walking, and 90 percent of the Land Bank ATM’s are within a kilometer from their next nearest neighboring backup (see Figure 1).
Figure 1: A two-dimensional histogram of the tessellation areas show that 68 percent of the areas have a coverage of 1.25 km2 or less.
The heavy concentration of branches arises in social and work congregation areas, like near universities, government offices, and business districts (see Figure 2). One such area is Quezon City where there is a prevalence of universities, like UP Diliman and Ateneo, as well as main government offices like the Land Transportation Office, the Social Security System, and the Department of Agriculture.
Figure 2: Smaller tessellation areas are found around business districts, like Manila City (left box), Quezon City (top right box), and areas near Makati Central Business District and Bonifacio Global City (lower right box).
Meanwhile, our results show that there are certain places—like north of Quezon City and southern Metro Manila—where one would have to travel farther just to reach an ATM. Coincidentally, these areas are relatively farther from business districts. The technique demonstrated in the study can be expanded to the entire country, and even be demonstrated using other facilities such as vaccination sites and gas stations to evaluate the accessibility of these commonly used facilities. ●
Elmo Domino R. Jose is an undergraduate Applied Physics student at UP Diliman. He may be reached at ejose@nip.upd.edu.ph.
This study, entitled “Evaluating geospatial efficiency of facilities through Dirichlet tessellation areas and centroid proximity analysis” appeared in the Proceedings of the 2021 Samahang Pisika ng Pilipinas held last October 20-22, 2021. It was co-authored by Damian N. Dailisan and May T. Lim. The authors of this paper are affiliated with the Complex Systems Team of the Instrumentation Physics Laboratory, National Institute of Physics.
The Philippine Collegian is looking for research papers to be featured in its monthly print issues. Submit a 500 to 700-word abstract, together with your bionote, to phkule.upd@up.edu.ph.
