ICACT20220062 Slide.22
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In tourism, autonomous electric vehicles may play an important role to provide ‘last-mile solutions’ facilitating multi-modality in moving tourists.
There are many ongoing studies which optimize operation cost about fleet sizing of transportation system based on queueing theory.
However, research to optimize initial investment is lacking.
Therefore, we suggest queueing theory-based investment minimization problem about introduction of the tourist transportation services with consideration of tourist’s waiting time constraint from the perspective of tour operators and investors. This research may bring proliferation of autonomous electric vehicles in tourism to the table.
Further, it could be applied to similar circumstance or system. Not only the tourism industry but also various industries effort to introduce AV and EV. Accommodations are participating in the EV industry by setting up charging poles. So, we will research on methods to optimize accommodations' investment cost and profit
Thank you for listening to my presentation.
|
|
ICACT20220062 Slide.21
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
This graph indicates investment cost obtained by multiplying A-EV’s unit investment cost by number of A-EVs. It's average tourist’s arrival rate and average A-EV’s service rate are specified as 100 requests per hour and 5 services per hour respectively. And, according to a curve which k is 2 in the figure 6, its charging station’s arrival rate is set as 0.688. Therefore, the number of A-EVs in the charging station is determined as 2. In this result, if tour operators want to provide transportation service under average waiting time of 5 minutes, they need to have 24 A-EVs at least. When the unit price of A-EV is unit I , their investment cost is calculated as multiplication unit I by 24.
|
|
ICACT20220062 Slide.20
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
this graph depicts the relationship between charging station’s arrival rate and the number of A-EVs in various number of charging poles with same charging rate. The arrival rate of each curve is enveloped in a range of below utilization 1 about given parameters. If A-EVs request charging frequently, the number of A-EVs in the charging station increase exponentially. When utilization of charging station closes to 1, almost every A-EV will be in the charging station. In summary, we can know that it is necessary to consider characteristic of tourist spot’s charging station to reserve the number of A-EVs for providing tourist transportation service within tourist’s waiting time threshold.
|
|
ICACT20220062 Slide.19
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
these show the busy probability that A-EVs exist at the charging station versus various arrival rates which in different charging service rates and different number of charging poles. On the curve of the charging station's service rate at 0.25 in above graph, the charging station is busy with an 80% chance at 0.2 arrival rate to the charging station with a charging pole. By the way in below graph, an additional charging pole can reduce this probability to 60% with the same arrival rate and same service rate. The better charging performance increases idle state probability which means no A-EVs in the charging station. A high charging rate may reduce investment cost of A-EV by enhancing service availability of A-EVs which provide tourist transportation service.
|
|
ICACT20220062 Slide.18
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Each tourist spot has their own area and tourist arrival rate. Also, service rate should be considered in various viewpoint such as service provider perspective and user perspective. Thus, to offer transportation service in tourist spot, parameters have to be determined properly.
Relationship between mean waiting time of tourists and utilization of service on static service rate is presented in this figure. The service rate is set as 1 and the arrival rate is set suitably according to the number of autonomous electric vehicles to envelope the utilization in between 0 to 1. These curves refer that mean waiting time increase exponentially with growing of utilization. Moreover, the mean waiting time can be reduced dramatically by arranging more autonomous electric vehicles. Also, when utilization approaches to 1, it proliferates. On analysis this graph in the frame of tourist spot operator's context, service provider can know minimum number of autonomous electric vehicles to operate transportation service while satisfying customers queueing time. Further, the service provider could obtain extra number of autonomous electric vehicles in order to cope peak arrival rate.
|
|
ICACT20220062 Slide.17
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
(14) stands for mean waiting time, which includes charging and waiting time.
On the same manner, the number of A-EVs in the charging station can be deduced like (15).
|
|
ICACT20220062 Slide.16
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Yield mean waiting time, using Whitt, 1976 and Medhi, 2003 approximation for the mean queueing time (Equation [13])
Each of two coefficient, 𝐶_𝑎^2 and 𝐶_𝑠^2 means, 𝐶_𝑎^2 means variation of the arrival time, which value is 1 (by Poisson Distribution) 𝐶_𝑠^2 means variation of the service time. This can be denoted as 𝜎_𝑠^2 ×〖(1/𝜇_𝑐^ )〗^(−2) , where 𝜎_𝑠^2 =𝑉[𝑡].
|
|
ICACT20220062 Slide.15
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
X, Y, Z denote the random variables charaEquation 𝐿_(𝑐, 𝑞) stands for average number of customers in the waiting line.
𝑃_(𝑐, 0) denotes the probability that there are no A-EVs in the charging station.
Mean waiting time in the queue(𝑊_(𝑐,𝑞)^(𝑀/𝑀/𝐾)) can be obtained with 𝐿_(𝑐,𝑞) and 𝜆_𝑐
|
|
ICACT20220062 Slide.14
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
X, Y, Z denote the random variables characterizing the battery capacity, the initial SoC and the requested SoC.
Variable 𝑐 is the threshold SoC value.
Every random variable included in equation (7) and (8) follows a truncated normal distribution.
|
|
ICACT20220062 Slide.13
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
According to charging station of reference model, charging station provide charging service to randomly arrival A-EVs in generic service time.
The expectation (E[t]) and variance (V[t]) of the charging time model are derived as equation (6), (7) respectively.
|
|
ICACT20220062 Slide.12
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
The optimization problem to minimize initial investment considering waiting time of service requester is formed as equation (1). Investment costs are determined by multiplying the number of autonomous electric vehicles by the unit price.
Unit investment of autonomous vehicle is noted ‘I_unit ’. Since the latency constraints represented by an ‘ε’ vary depending on the characteristics of objective system, the suitable values for the system must be considered.
the customer’s waiting time, W_q, can be estimated through equation (2) to (5).
By solving this optimization problem with corresponding arrival rate and service rate of targeting circumstance, a relationship between the number of service autonomous vehicle and waiting time of service requester is obtained. This relationship shows exponential decrease of customer’s waiting time while the number of autonomous electric vehicles increase linearly.
|
|
ICACT20220062 Slide.11
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In the same manner, this figure is analyzed follows.
|
|
ICACT20220062 Slide.10
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Mapping relationship of target environment with queueing theory is presented in this figure.
Applying queueing theory notations to our target environment, average arrival rate is considered as average rate of requesting transportation service per time period.
Average rate of transportation completion to required destination during a time period is considered as average service rate.
Average time spent waiting the service autonomous electric vehicle is considered as mean waiting time.
The number of service autonomous electric vehicles for providing transportation service is considered as the number of servers.
|
|
ICACT20220062 Slide.09
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
this figure shows our target environment. This tourist spot consist many places to tour like temple, amusement park, pond, restaurant, souvenir store. and it provides transportation service on tourist’ or a group of tourists’ demand with autonomous electric vehicles. The tourists can visit specific place or walk around in the tourist spot. Also, they can use a service providing autonomous electric vehicles by calling it if they need, and they can get off from the autonomous electric vehicles at anywhere in tourist spot.
|
|
ICACT20220062 Slide.08
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Each tourist spot has its own number of visitors and its own tour length.
So, when tour operators consider introducing tourist transportation services, it is necessary to take into account the number of autonomous electric vehicles suitable for their tourist spot.
hence, we suggest queueing theory-based investment minimization problem about adopting tourist transportation services with consideration of a waiting time constraint from the perspective of tour operators and investors. tour operators and investors can see how many autonomous electric vehicles are required at least in their circumstances by applying this optimization problem.
|
|
ICACT20220062 Slide.07
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
From a tourist's point of view, these tourist transportation services could increase access to tourist spot and provide an improved travel experience by giving travel convenience to tourist.
Consequentially, more tourists could be attracted to tourist spot.
However, from the perspective of tour operators and investors who want to introduce tourist transportation services, it is necessary to consider number of A-EVs because they have limited capital and investment.
For example, if number of A-EVs is too small, it requires less investment, but it reduces service satisfaction by increasing waiting time to use.
Besides, the waiting time in these services causes inconvenience to users.
Contrariwise, if number of A-EVs is too large, it reduces waiting time and increases service satisfaction, but it requires a large investment cost.
According to a study conducted on wait perception, people feel that unoccupied time is longer than occupied time.
Additionally, it should be noted that the A-EV cannot provide services while the A-EVs waits at the charging station for charging
Thus, they need to consider the waiting time of users, who want to be serviced.
|
|
ICACT20220062 Slide.06
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In most tourist attractions, transportation services using the autonomous vehicles have been demonstrating to help tourist's travel.
For example, in Chatan Town, Okinawa prefecture of Japan, for the purpose of achieving driverless autonomous driving mobility services, two ministry worked together to advance a demonstration test for technologies and services involving autonomous driving vehicles, making use of a remote-controlled system.
On March 31, 2021, They shown that driverless autonomous driving mobility services on a seaside course.
|
|
ICACT20220062 Slide.05
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
In tourism, autonomous vehicles are anticipated to provide ‘last-mile solutions’ facilitating multi-modality,
therefore that may play an important role in moving tourists. In terms of cost, because autonomous vehicles have no driver, the cost of travel could be substantially reduced.
On the other hand, from a convenience point of view the autonomous vehicles could alleviate the inconvenience of renting a vehicle in a foreign environment. Also, that could minimize risk of jet lag, fatigue, misunderstanding of new driving rules, and cultures of mobility.
|
|
ICACT20220062 Slide.04
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Autonomous driving and electric vehicles are very closely related. Many studies have predicted that electric vehicles will bring about a groundbreaking environmental pollution reduction effect.
|
|
ICACT20220062 Slide.03
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
We are witnessing an evolution into the automotive industry in transition period from human driven vehicle to autonomous vehicle.
In 2014, a classification document was published by Society of Automotive Engineers International. This document describes 6 different driving levels according to 6 types of driving scenario with characteristic driving function requirements.
As stated in this document, autonomous vehicles have been progressing gradually . This advance in autonomous vehicle industry is prospected to affect in whole industry.
|
|
ICACT20220062 Slide.02
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
these are contents of our presentation
|
|
ICACT20220062 Slide.01
[Big slide for presentation]
|
Chrome Text-to-Speach Click!! |
 |
Hi everyone, I am LaeYeop Lee.
I'm undergoing master course in Chungbuk national University.
My advisor is professor Seong Gon Choi
This video presents Investment Cost Minimization of Autonomous-Electric Vehicles based on Queueing Model
|
