Allocation planning for probe taxi devices based on information reliability

•We analyze the efficient allocation of probe devices to taxi dispatch centers.•Pass frequency and required sample size are modeled from actual data.•Efficiency is defined as the probability of satisfying the required sample size.•Efficiency of data collection changes according to the data collection period.•Efficient data collection requires the allocation of probe devices to suburban areas.

In recent years, the gathering of information about road traffic conditions using probe vehicles has begun in many cities. The travel time obtained using probe systems varies among vehicles because of driver characteristics and/or GPS errors even when traffic conditions are constant. To generate traffic information with high reliability, a relatively large number of samples is needed. Also, for efficient information collection, it is necessary to give careful consideration to the allocation of probe vehicles. In this study, the authors analyse the efficient allocation of probe vehicle devices to taxi dispatch centers (TDCs) by using taxi-probe data collected around Nagoya City in Japan. The results show that the efficiency of data collection changes according to the total number of probe devices and the data collection period. In general, efficient data collection requires the allocation of many devices to suburban areas surrounding the central urban area.