In this study, the changes in urban traffic by autonomous driving technology are summarized. In particular, changes in parking behavior and parking facilities, which are most sensitive to changes from AV, are analysed. When autonomous vehicles are parked in an existing parking facility, the capacity of that parking facility increases by 30 ~ 60%. An AV exclusive parking garage can lead to a further 70% land use efficiency. In particular, the combination of a mechanical parking lot and autonomous driving technology can increase parking capacity by more than 75% and the land utilization rate by more than 14%. 
Autonomous vehicles avoid crowded and expensive urban parking lots by sending deadhead back to home, or moving to nearby cheap parking lots. In an opinion poll about the usage of autonomous vehicles, 83.7% of existing passenger car drivers responded that they would switch to autonomous vehicles. On the other hand, 58.2% of users of public transportation responded that they would switch to autonomous vehicles. As a result of a poll related to the selection of a parking lot when going to the city center in an autonomous driving car, the choice of a parking lot outside the city center was the most common among passenger car users and public transportation users. It was also found that 21 ~ 27% of respondents chose a parking lot in the city center, and that 8 ~ 12% answered that they returned their car to their home. Due to the demand for autonomous car parking, an additional VKT (Vehicle · km) of up to 4.7 million vehicles / km occurred, which is about 30.0% of the total driving distance of Seoul city (53,426,401 vehicles / km in 2014). In scenarios where a single fare is charged at a rate lower than the city center, a VKT of about 18.9 million vehicles / km, which is about 35% of the total driving distance of Seoul city, is added, and there is an increased travel distance of 31.9 km per vehicle as well. 
Therefore, the management scope and characteristics of the existing traffic demand management policy were diagnosed in order to envision a demand management policy that reflects the characteristics of autonomous driving cars. This study proposes a demand management zoning system that selectively applies existing traffic management policies that are appropriate for each region’s traffic characteristics.
The principle of “Zoning” for traffic demand management considering autonomous vehicles’ characteristics is as follows. As one enters the city center, it adopts a strong demand management policy and encourages the transfer from autonomous vehicles to public transportation in a variety of ways, including strengthening transfer facilities between modes. Area 1 of Jongno-gu, Jung-gu, Yeongdeungpo, and Jongno-gu, which are the targets of Zone 1, will be designated as special management zones, and a strong demand management policy will be implemented therein to limit vehicle entry by congestion tolls. In Zone 2, the “Kiss & Ride” method is applied to reduce the demand for autonomous vehicles to enter the city center. Autonomous vehicles drop passengers off and return to the low-cost parking lot or back home, and passengers then use public transportation. Zone 3 is the area coping with the demand of the wide-area traffic demand. Autonomous driving cars park in AV exclusive parking lot with low rates at the outskirts of the city, and drivers then use public transportation to their destination.