▶The MOIS will use AI-learning data to optimize deployment and waiting spot/depot allocation of mobility assistance vehicles for the transportation-disadvantaged.
▶The analysis is expected to help the city lower the travel distance of assistance vehicles by 41% and greatly reduced users’ waiting time.
▶Daejeon plans to apply the data starting in October, and the ministry plans to standardize and deploy the model across Korea.
On October 17 (Tuesday), the Ministry of the Interior and Safety (Minister: Lee Sang-min, MOIS) announced the completion of the Artificial Intelligence (AI)-Based Analysis on the Deployment Optimization of Support Vehicles for the Transport-Disadvantaged.
The MOIS Integrated Data Analysis Center and Daejeon Transportation Corporation (DJTC) developed the analysis model based on the data from the city of Daejeon between July and early October.
The city government currently operates 96 mobility assistance vehicles for people experiencing difficulties accessing transportation services. In 2022, the vehicles were deployed 1.10 million times in total. When a transportation-disadvantaged person requests deployment by contacting the call center, the nearest driver from the caller is assigned.
However, the waiting spot of each vehicle has been arbitrarily selected by the driver without considering where the vehicles are needed. As a result, callers had to wait for an average of 22 minutes (up to 51 minutes) to board the vehicle, causing significant disruption to their mobility.
The depots for assistance vehicles were also allocated at the local government’s convenience without considering the service users’ locations, which added to their inconvenience.
The MOIS and DJTC addressed the issue by analyzing the use of assistance vehicles for transportation-disadvantaged persons by time of day and identifying the number of departing areas and destinations with the highest number of users. Based on the findings, they identified five towns (or dong (洞)) as potential waiting spots for minimizing users’ waiting time.
They also used AI to assess the locations of the existing seven vehicle depots, and suggested six new depot locations to minimize deployment time.
Deploying vehicles from the newly proposed depots have been confirmed to reduce travel distance by 41%, from 5.0km on average to 2.9km.
The city government and DJTC plan to use the analysis results to change the waiting spots for the mobility assistance vehicles starting in late October and gradually relocate vehicle depots starting next year.
The MOIS expects that the analysis data will greatly improve the mobility of the transportation-disadvantaged.
As such, the ministry plans to standardize and deploy the model across Korea so that other local governments operating mobility assistance vehicles can use the data.
Source : Integrated Data Analysis Center, MOIS