Meet Dakota Pekerti (middle), a Master student from the University of California, Berkeley, who has successfully completed his AIFIS–CAORC sponsored research on planning electric motorbike infrastructure in Jakarta.
His research explores how to identify optimal locations for battery swapping stations (BSS) by first understanding how motorbike riders actually move through the city. Drawing on demographic data and travel diaries collected from riders in Jakarta, Dakota applies statistical methods—including Principal Component Analysis (PCA) and k-means clustering—to identify distinct rider groups. These clusters help capture different travel behaviors and allow him to simulate how each group would use electric two-wheel vehicles (2WEVs), including when and where they would need to recharge.
To bring these insights together, Dakota developed an agent-based model (ABM) that integrates real-world rider behavior with the technical characteristics of electric motorbikes. This approach enables the simulation of individualized travel patterns, offering a more grounded alternative to traditional infrastructure planning methods that often rely on aggregate data alone. Through this model, the research highlights where demand for battery swapping stations is likely to be highest and how infrastructure can better align with everyday mobility patterns.
This modeling approach marks a shift from conventional infrastructure planning methods, which often prioritize efficiency metrics like cost minimization or distance coverage. Instead, the ABM emphasizes lived experience, simulating the daily decisions of riders and translating them into patterns of energy demand. The result is a more nuanced understanding of where infrastructure is truly needed—sometimes revealing unexpected areas of high demand that would be overlooked by traditional models.
Like many field-based projects, the research process came with challenges. Plans for in-person interviews in Jakarta were affected by periods of unrest and administrative delays, requiring a shift to remote interviews after returning to Berkeley. Additionally, the initial plan to expand the study to other cities, such as Medan and Makassar, had to be scaled back due to time and data constraints. Despite these hurdles, the Jakarta-focused model was successfully completed and demonstrates strong potential for informing real-world infrastructure planning.
Collaboration was central to the project’s success. Partnerships with local institutions, including PPM School of Management, and survey support from Jakpat enabled effective data collection and refinement. These collaborations also helped strengthen academic connections between Indonesia and the United States.
Ultimately, this study offers more than a technical solution—it presents a people-centered framework for Indonesia’s electric mobility transition. By grounding infrastructure planning in the lived experiences of riders, it points toward a more equitable and responsive approach to urban transformation.
Congratulations to Dakota Pekerti on the successful completion of his fieldwork!