01-11-2016, 02:41 PM
1463041243-R6IJ3QC.pdf (Size: 525.5 KB / Downloads: 92)
ABSTRACT
Mobility, context-awareness and payment combined provide a
customer with a completely new setting of consuming services at
any time and any place. We introduce SmartRestaurant service,
which allows customers to use mobile devices for ordering and
paying lunches from a nearby campus restaurant beforehand.
Further, SmartRestaurant provides the restaurant with means of
adjusting the sales with production capacity and prior knowledge
of upcoming orders. We present a user evaluation of the system in
form of a field trial in the real environment of use.
INTRODUCTION
In recent times, there has been dramatic penetration of powerful
mobile devices with ability to enhance the mobile environment. A
number of services have been deployed for mobile terminals
including news, directory services and payment services.
If we are paying or selling goods through a fixed or a wireless
network, we can consider that as eCommerce (electronic
commerce) or mCommerce (mobile commerce). ECommerce like
mCommerce can be classified to indirect and direct (tangible and
intangible) business [3], depending on type of the goods that are
being sold via the fixed or wireless network.
Mobile services are available at any time and any place, which has
made the mobile phone technology so popular. Usually the chargeable added value mobile services provide ringing tones,
icons and other intangible digital goods, which are being
downloaded to the customer’s device. Shopping via mobile
devices is projected to be a major business channel in the coming
years.
Mobile payment (and also mobile payment systems) can be
classified to remote and locale payments. The difference is
obvious: in remote payments mobile network is used to route the
transaction information between the payment system and user
device as in local payments the user connects to the POS (Point of
Sale) terminal physically or with a short range access protocol like
Bluetooth or infrared. Both remote and locale payments systems
are being used in various applications.
The remote payment systems are usually used in services that do
not require handling of physical goods. In these content services
the user’s location is irrelevant. For example Sonera, the leading
operator in Finland, provides several mobile services for its
subscribers. Users can order different kind of SMS (Short
Message Service) newsletters and digital content to their mobile
phones [13].
In 2002 Sonera launched a mobile payment service Sonera
Shopper [14]. It is an online mobile phone service, which is being
used with SMS-messages. The mobile user needs to register to the
service before the use. With the Sonera Shopper mobile payment
method the users can pay their purchases in several stores in
Helsinki and also parking fees in couple of cities in Finland. The
payments are being made using the user’s Sonera Shopper
account or the user’s credit card.
ZonePay [19] develops mobile ordering systems and payment
solutions. For the time being their major product is the iWaitLess
service. The service is directed to ‘take away’ restaurants and it
can be used via mobile device or a web-browser. The customers
can order and pay servings before hand from the restaurants,
which are signed-up to the service. The service has been tested
and deployed in the United States.
The local payment systems are used in user’s proximity
environment. The distance between the POS and the user depends
on the technology that is used. Nokia and 2Scoot have tested
RFID technology as a local payment system [9].
In mobile domain the service system can benefit from
understanding the usage context of the users by automatically
adapt the service to fit users’ needs. The automated assistance is
especially important with limited user interface devices like
mobile phones. Generally, the context information can be
understood as any information that can be used to describe the situation of a person, place or object that has significance to the
interaction between user and the service [2]. Useful context
information exists in various levels of abstraction. Some examples
of contexts are the location of the user, surrounding weather,
user’s current yearn, social relations with nearby users, bandwidth
of the user’s mobile device, screen size of the mobile device, etc.
This information is eventually used to provide the user more
valuable services like personalized news, guidance services,
context-based directory services and other “smart” services.
In the previous mobile payment concepts context-awareness is not
needed, since the user’s location and contextual information do
not provide any added value for the service. VISA presents a
scenario of a traveling business executive who decides to get
something to eat [18]. The person orders and pays a meal from a
local restaurant using his wireless handheld computer. Using a
carrier that calculates the customer’s proximity to the restaurant
and provides directions to the restaurant the customer gets his
meal eventually. The paper does not present any actual
implementation for realizing the scenario, but it just gives an idea
how context-awareness could be used to provision more valuable
services. Other related work on context-aware mobile commerce
includes the context-aware and location-based mobile ecommerce
server by Jin and Miyazawa [7] and Varshney’s
conceptual study of location management in mobile commerce
applications [17]. Both papers remain somewhat theoretical,
however, since no concrete deployment and its evaluation are
reported.
The novel contribution of this paper is the deployment and
evaluation of the SmartRestaurant service in the true environment
of use with genuine end users. SmartRestaurant is designed for a
context-aware mobile service system such as SmartRotuaari [11],
allowing customers to use mobile devices for ordering and paying
lunches from a nearby restaurant beforehand. Further,
SmartRestaurant provides the restaurant with means of adjusting
the sales with production capacity and prior knowledge of
upcoming orders. An empirical evaluation is carried out in the real
environment of use in a form of field trial involving one restaurant
and 20 test users. The SmartRestaurant service is a new feature in
the SmartRotuaari service system, which offers context aware
mobile multimedia services to consumers at a city center [11].
One of the interesting features in the SmartRestaurant is the
ability to entice the user to make the lunch order and payment
from a distance to the restaurant. This behavior could be further
stimulated by providing the user with the means and information
needed for the purchase in a simple and intuitive way. This can be
achieved with personalized mobile advertisements pushed to the
user’s mobile device. Kaasinen [8] has found out that the users’
attitude towards push type mobile advertising is positive if the
information is relevant. An implementation and empirical
evaluation of a location-aware mobile advertisement system based
on WAP Push is presented by Aalto et al. [1].
Context-awareness of the presented deployment of
SmartRestaurant is limited to the implicit incorporation of the
end-user’s context when (s)he is placing the order. However,
context-awareness could be enhanced in a straightforward manner
by integrating SmartRestaurant seamlessly into a context-aware
architecture such as SmartRotuaari [11], or by incorporating push
type notifications using the aforementioned location-aware mobile
advertisement system [1].
This paper is organized as follows. The SmartRestaurant system is
introduced in Section 2. An empirical evaluation of the system is
described in Section 3. Section 4 provides concludes the paper
with a discussion on various aspects of the system and future
work.
2. SMARTRESTAURANT SERVICE
SmartRestaurant is a web service for mobile users that has been
designed to enhance a restaurant’s production and delivery
process. The SmartRestaurant actors are categorized to customers
and employees.
The customers (also referred to as end-users) are normal
customers except they use the SmartRestaurant to order and pay
their lunch beforehand. They can browse the SmartRestaurant’s
menu with a mobile device, order and pay one or more meals, and
schedule the delivery time of their order relative to their current
context (time, location) so that the food will be hot and fresh
when they enter the restaurant.
The employees of the restaurant configure the service and prepare
the ordered meals. SmartRestaurant allows the restaurant to
automatically adjust the sales with the production capacity.
SmartRestaurant also provides the restaurant with a prior
knowledge of upcoming orders.
2.1 Distributed architecture
SmartRestaurant’s distributed architecture is illustrated in Fig. 1.
Although the illustration contains references to specific local
resources utilized in the field trial, the architecture is built of
standard Internet components, which allows a robust deployment.
1. SmartRestaurant web service is installed on a host (named
rotuaari.net in our case), which is connected to the Internet via a
firewall.
2. The payment service is installed on a host, which is connected
to the Internet via a firewall. In our system we use the e-maksu
payment service [4] hosted by PPO, a local operator. The e-maksu
service provides customers with accounts on which money is first
transferred from a regular bank account. The money on the
customer’s e-maksu account can then be used for paying for
content and services of service providers having subscribed to the
e-maksu payment service. The payments are aggregated to the
providers’ e-maksu accounts and then transferred to their bank
accounts.
3. Mobile users can use the SmartRestaurant with smartphones,
which provide means for Internet access, e.g. GPRS (General
Packet Radio Service) connectivity, and a web browser capable of
showing XHTML pages. In the field trial phones used the GPRS
network of Octopus, which is a local development and testing
environment for new, innovative mobile applications and services
[10].
4. Mobile users can use the SmartRestaurant also with a PDA
(Personal Digital Assistant), which features wireless connectivity
and a XHTML capable web browser. In the field trial PDA’s
connected to the Internet via panOULU, which is a local public
access network comprising of a number of WLAN hotspots [12].
PanOULU provides wireless Internet access to accounted users,
which are authenticated by an access controller.
5. Restaurant employees operate the SmartRestaurant service with
a laptop equipped with WLAN connectivity and a standard web browser. Laptops with wireless connectivity can be placed freely
at the restaurant premises and they also allow the employees be
mobile.
Restaurant employee interface
The restaurant employees access the SmartRestaurant service via a
web browser. After logging in with an employee account the
employee is provided with the main menu, which provides access
to following functions: configuration of the service, status page of
orders (used by chef), listing of all orders (used by cashier) and
reports (used by cashier).
Fig. 4(a) shows the configuration page, which allows for
specifying the daily menu (meals and their prices) and the
attributes of daily production process (opening hours, serving
hours, delivery period, preparation time of a single order, and the
maximum number of orders per delivery period). Based on these
attributes, the SmartRestaurant automatically adjusts the sales
with the production capacity, by keeping a record of booked
delivery times and offering only available delivery times for new
orders. In the screenshot the attributes are set so that at most 10
orders can be sold per a single delivery period of 15 minutes.
Fig. 4(b) shows the status page of orders, which is used by the
chef in the kitchen. When a new order arrives, the chef is alerted
with both an audible (randomly chosen from a set of audio clips,
can be switched off) and a visible notification. A new entry
displayed in green appears on the status page, showing detailed
information of the order (delivery time, number of portions, type
of meal). Having prepared the order the chef taps the “prepared”
button on the corresponding row, and the entry turns into yellow.
EMPIRICAL EVALUATION
We evaluated the SmartRestaurant service in form of a field trial
in the true environment of use involving one restaurant and 20 test
users.
The operative goal of this field trial is to achieve results
concerning end-users reactions with chargeable content due a
mobile terminal in a context-aware environment, where the endusers
evaluate their own time and location.
3.1 Setup of the field trial
The field trial took place at the campus of the University of Oulu.
SmartRestaurant service was offered by restaurant Kastari, which
is one of the many campus restaurants operated by Uniresta Ltd.
Kastari is a lunch restaurant offering two principal types of meals,
a lunch buffet collected by the customers and a daily special lunch
ordered and prepared individually on order-by-order basis. When
the customer orders the special lunch, (s)he gets an order number
and takes a seat in the lunchroom, waiting for her/his order to be
prepared and served. Kastari is operated by three full-time
employees, a chef and two cashiers working in turns in the kitchen
and at the cash register.
Given the service model of Kastari, the SmartRestaurant service
was configured to offer the daily special lunches. With
SmartRestaurant customers were able to order and pay their
special lunch beforehand, thus avoiding the need to queue up at
the cash register. Further, customers could specify the preferred
delivery time, avoiding the waiting in the lunchroom.
The restaurant employees operated the SmartRestaurant with three
computers. A laptop equipped with WLAN connectivity and a
large touch screen was placed in the kitchen for the chef. Another
laptop with WLAN connectivity was placed next to the cash
register for the cashier. The employees configured the service with
a standard desktop PC available in the office,
The chef configured the SmartRestaurant service so that at most
10 lunches could be sold for a single delivery period of 15
minutes, first period of the day scheduled for 30 minutes before
Kastari was opened. The maximum number of portions per order
was set to five. The SmartRestaurant service was set to
automatically close at the same time as Kastari was closed.
The restaurant employees were trained to operate the
SmartRestaurant service and the new delivery process was
rehearsed before the start of the field trial.
3.2 Test users
In total 20 test users participated in the field trial, 15 males and 5
females. All test users were Finnish, due to recruiting being
carried out in Finnish. The majority of the test users were in their
twenties, 8 (40 %) were 18-24 and 11 (55 %) 25-34 years old.
There was only one person (5 %) in 35-49 age group. 11 (55 %)
test users were students and the remaining nine worked in the
University of Oulu.
18 test users reported to visit the Kastari restaurant regularly, at
least once per week. One person (5 %) used to visit the restaurant
only once per month and one had never visited the Kastari before
the field trial. Only six persons (30 %) informed that they usually
eat the special lunch at least once per week. Four persons had
never tried the special lunch at Kastari.
All test users owned a mobile phone, while only four (20%) test
users had their own PDA. Test users were frequent mobile phone
users, for only seven people informed that they use SMS
messaging only one or two times a week or less often. Other
mobile services were used less often, as illustrated in Fig. 6.
18 test users reported to pay over 75 % of their bills with Internet
banking. However, only two (10 %) of the test users had paid bills
via a mobile phone.
Some of the test users did not consider downloading logos and
ringing tones as mobile payment and were surprised to hear that
downloading digital content to their mobile phone is classified as
such. This reflects the transparency of the post-payment system
employed by network operators in contrast to e.g. the e-maksu
system utilized in SmartRestaurant, where the customer has to
deposit money to her/his account beforehand.
Almost every test user (18 out of 20, 90 %) told to know the
difference between GPRS, LAN and WLAN network
technologies.