Page 151 - Proceeding of Atrans Young Researcher's Forum 2019_Neat

“Transportation for A Better Life:
Smart Mobility for Now and Then”

23 August 2019, Bangkok, Thailand

fold improvement in parcel delivery efficiency. customers reach their collection point, enabling high
Other benefits include reduced traffic congestion service level and “instant pickups”. The duration
and carbon emissions from fewer multiple-location between sellers dropping off the orders and buyers
small parcel deliveries.” 19 picking up the orders should be minimized through
AI. This optimized scheduling is what enables the
The use of lockers as last mile delivery is reduction of operational costs of collection point
very promising. However, for successful operation, rentals. After a period of service, predictive analytics
logistic companies need to ensure efficient locker can be done to forecast demand and increase
utilization and turnover. Locker installation or rental optimization as well.
costs are high, especially in strategic and populated
locations like interchanges. To maximize profits, For buyers, this solution minimizes
logistic companies would need to maximize locker uncertainty and wait time associated with normal
utilization rate. Late collection hinders locker home deliveries or queueing up at stores. With AI
turnover and reduce efficiency. recommending optimal time for preparation and
delivery, sellers are also able to better plan their
3. Methods time, reducing uncertainty and confusion that occurs
This study explored an integrated concept of during peak hours as well.
e-commerce and public transit for commuter and
delivery providers. While travelling, customers are Self-collection delivery methods increase
able to shop-on-the-go, order through a mobile app the efficiency of last mile delivery as there are
and collect it at a parcel collection lockers situated consolidation of orders, fewer number of delivery
along their commuting route. This concept was points and reduced number of point-to-point trips by
developed based on the concept of “What I want, using public transport network. This has the
where I want, when I want it”, with a goal to increase potential to significantly reduce both fixed cost
the flexibility and predictability of online delivery (manpower logistics and fleet) and variable costs
for a more customer-oriented service. It is focused (fuel) for the delivery companies. Meanwhile, self-
on delivering products that consumers want to the collection stations can be set up at strategic locations
location which is convenient for them, at the time like train stations or interchanges. In order for a
that is convenient for them by taking advantage of seamless collection process, lockers can be installed
public transportation modes instead of being with biometrics authentication systems. For
constrained by fleet size and manpower/labour. biometric collection points, customers’ biometrics
data will have to be collected prior to self-collection.
The key enabling factor for this concept is in This can be done through the uploading of a picture
combining pervasiveness of public transport in via the mobile app for face recognition systems. For
Singapore to achieve just-in-time delivery at self- non-biometric collections, the generation of QR
service lockers and to optimise their turnover rate. code for collection will suffice.
This is achieved by using real-time tracking of
commuters from Mobility-as-a-Service (MaaS) In order to better understand current trends
mobile app, which offer transport multimodal and evaluate the concept, quantitative and qualitative
routing combined with e-commerce purchasing and market studies were carried out. Firstly, an online
payment ability. This enable customers to shop and survey was conducted with 71 respondents from
make online payment while travelling from one Singapore to understand the current mobile
destination to the next. By processing real time commerce trends, challenges and gaps in Singapore.
location of customers, delivery couriers, and Then, an interactive co-creation group discussion
scheduled public transport operation - an optimal was conducted with 77 respondents from Singapore
and predictable delivery collection time can be to understand their views on the solution concept and
determined and coordinated through artificial potential opportunities and challenges. Survey
intelligence (AI) optimization algorithms. methodology studies the sampling of individual
units from a population using the associated survey
With this optimal collection time, sellers are data collection techniques, such as questionnaire
able to know exactly what time buyers will reach the construction and methods for improving the number
20
collection point and prepare accordingly. This helps and accuracy of responses to surveys. This
to ensure that products are made to be ready when methodology provide a means to explore the

126