Evaluating Delphi survey accuracy in transportation: Evidence from Japanese technology foresight

Technological Forecasting and Social Change. 2026. Vol. 224. Article 124496.

Niyazov S., Maibakh O., Alexei Sukharev, Kulakova T., Ufimtsev A.

This research evaluates the accuracy of Delphi surveys in technology foresight with a focus on the transportation sector by determining the realization status of topics forecasted in the three Science and Technology Forecast Surveys administered by Japan's National Institute of Science and Technology Policy (NISTEP) in 1992, 1997, and 2001. The analysis included 167 topics, representing 647 observations when different Delphi rounds, survey iterations, and respondent pools are taken into account. Through expert verification, an overall forecasting accuracy rate of 25 % was found.

The results highlight a “forecastability pit”: topics in early and late technological stages demonstrate higher accuracy compared to middle stages, when technologies first enter practical use or are improved. Accuracy was lowest for “aviation” (11 %), “railway” (19 %), and “road” (25 %) sectors and highest for “marine” (31 %) and “other new transportation sectors” (36 %). Greater accuracy was not found to be correlated with higher levels of self-rated expertise, second-round Delphi forecasts, a topic's perceived importance and anticipated magnitude of effects.

The low accuracy rate may stem from the regulatory and technological complexities inherent to the sector. Overall, the study underscores the need for methodological refinements and a reassessment of the role of expert judgment processes in S&T Delphi.

Keywords: Foresight, Delphi, Дельфи, Transportation, Forecasting accuracy, Expert methods, NISTEP, Форсайт, Точность прогнозирования, Экспертные методы, Транспорт, NISTEP