TOWARD PERVASIVE COMPUTING IN RESTAURANT

Yang Xiang, Wanlei Zhou, Morshed Chowdhury

School of Information Technology, Deakin University, 221 Burwood Hwy, Burwood 3125 VIC, Australia

Keywords: Pervasive Computing, Web-based System, Ordering System, Wireless Network

Abstract: In this paper, an example of pervasive computing in restaurant, a wireless web-based ordering system is

presented. By using mobile devices such as Personal Digital Assistants (PDA) and WebPad, customers can

get many benefits when making orders in restaurants. With this system, customers get faster and better

services, restaurant staff cooperate more efficiently with less working mistakes, and enterprise owners thus

receive more business profits. This system has multi-tiered web-based system architecture with good

integration and scalability features, and is client device operating system fully independent. Details of

design and implementation of this system are presented.

1 INTRODUCTION

In order to improve productivity and operational

efficiency, today more and more enterprises pour

their investments in mobile solutions. The

technology of mobile computing today is matured.

When mobile computing devices are available

throughout the physical environment, finally they

become effectively invisible to the users (Weiser,

1993) because they are so familiar with them that

they just ignore computing devices’ existence.

Obviously the conduct of business and services

over portable, wireless devices has lots of

advantages. The miniature size of mobile terminals

and the fact that they have light weight and can be

carried everywhere makes them an ideal channel for

offering personalized and localized services. Some

companies developed wireless solutions for

inventory management, Point-of-Sale (POS), traffic

management and so on (Stanford, 2003). Their

approaches are all based on wireless Client/Server

architecture, which rely largely on computing

capability of mobile devices. Therefore they face the

problem of scalability and compatibility, although in

small scale application, it perhaps appears not a

serious problem.

To build wireless applications, web-based

information access technology is a fairly good

choice (Catarci, 1999). By web-based technology,

we can develop a friendly and easy-to-use system

with least administrative work. In this paper, we

present a wireless web-based ordering system,

named iMenu, which has advantages of flexible

functions, easy to operate, and easy to learn. It saves

a restaurant’s human resource and improves working

efficiency. We find our system is one of the first

approaches that use wireless multi-tiered Internet

architecture to build commercial pervasive

computing systems in the restaurant industry. It does

not need a powerful mobile client in a pervasive

computing environment, while keeping a modest

thin client has many benefits (Satyanarayanan,

2001), as shown later.

2 ORDERING IN A RESTAURANT

The business process of ordering in a restaurant is

not complicated. However, in a large size restaurant

it takes tens or even hundreds of people to work

together to serve a single customer. The first step is

that customers browse printed menu, choose their

favourite foods, and then tell waiters what they

want. At the same time, waiters are busy writing

what customers order on an order pad. After finish

ordering, waiters transfer the paper of ordering to

kitchen or cashier (later cashier will pass it to

kitchen). When chefs in kitchen receive orders, they

check if there are enough materials for cooking from

stock staff. If items are run out, they inform waiters.

Afterwards, waiters may apologize for sold out

items and customers may make new orders. If the

food is ready, chefs transmit it to waiters for serving,

or, in large size restaurant, there are special workers

who do this transmitting job from kitchen to waiters.

At the meantime, waiters check the ready foods on

312

Xiang Y., Zhou W. and Chowdhury M. (2004).

TOWARD PERVASIVE COMPUTING IN RESTAURANT.

In Proceedings of the First International Conference on E-Business and Telecommunication Networks, pages 312-317

DOI: 10.5220/0001382503120317

 SciTePress

the copy of order list. After finish their dinner,

customers make payment with waiters’ help.

Restaurant managers then read business report,

arrange human resource, and handle accidental

situations.

Here we only describe the simplified business

process of ordering in a restaurant. In real condition,

some orders may be transferred to not only kitchens

but also bars, which provide wines and so on. From

the analysis we can see waiters are under high

pressure to provide good services because they have

too much work to do and too many places to run.

And all the information is recorded on paper,

mistakes may be easily generated, say nothing of

disaster when one piece of these papers is lost during

transmission. If customers are not satisfied with

services, they may ask for changing or returning

foods. It is not always restaurant’s fault, but this

situation is a real case. So waiters, chefs, cashiers

have to change all things according to the changed

order. Generally, the business flow is quite simple;

however, to accomplish all these tasks is

burdensome for both the customers’ side and the

restaurant side without an efficient ordering system.

Our initial thinking to develop this wireless

web-based order system, iMenu, is to minimize the

work of restaurant staff and to eliminate working

error as much as possible. Since most of restaurants

have POS system today, it should be able to access

previous system’s database. Let’s see how iMenu

works as follows. Customers browse menu, make

orders, enjoy their foods and pay as before. The only

changed customers can feel is that the response time

is shortened and error is reduced. The waiter hold a

mobile device (PDA or WebPad), which is easy to

carry, and a pen to tap the small glimmer screen.

With the handwriting recognition technology and

soft keyboard, waiters can input words easily. By

using our system, waiters can record customers’

order even faster than traditional tools such as pen

and paper. When ordering is finished, the waiter

strokes a button; all information is transferred to

kitchen, stock room and cashier. If some items are

run out, a small alert message box is shown on

screen of mobile device in a second to let waiters

and customers know. To change an order is also a

simple job. Waiters just need to recall the order,

modify and transmit it again. Chefs, stock staff and

cashiers can know everything clearly on screens in

their rooms. As an alternative of output device in

kitchen, we also use the printer. It may be better for

chefs to get information from printers because it is

easy to take notes or comments on papers when

needed. And computer monitors don’t work well for

a long time in a relative high temperature and dirty

atmosphere environment. Without walking from

place to place, talking in noise, waiters, chefs, stock

staff, and cashiers can work together in an easy and

calm manner. iMenu offers clear benefits for

restaurant’s customers, restaurant managers and

servers. It minimizes the human interaction with

information systems.

3 IMENU SYSTEM

3.1 Requirement Analysis

Every category of restaurant staff has its own

requirements. Restaurant managers are eager for

more business profits, less human resource and easy

management of both restaurant assets and staff.

They want to compress cost as much as they can. So,

reducing staff’s working error by iMenu becomes

the key issue to get better inventory management

and less waste. All items in inventory should be

stored in a database, which keeps management in an

orderly way. On the other hand, human resource

management is also the same way. Managers should

be able to fetch any staff’s information and track

their working time by iMenu’s login system.

Moreover, as an example of “calm technology”,

managers could send instruction to staff without

facing them, taking benefits of wireless system.

Therefore, iMenu should have built-in instant

message system to let managers communicate with

their staff, which is a trivial work they must do

before.

Chefs need fast and accurate orders. Printers in

kitchen give them many benefits. They just tick the

items on papers when they are finished. Touch

screen is also helpful when they need input message,

for example, when items are run out. In a word,

iMenu system should save their time and help them

avoiding mistakes.

For waiters, they need an easy-to-use ordering

system that helps them minimize working error.

First, at any time they will know which tables are

empty and which ones are being occupied. iMenu

needs a real-time table map to show table status.

Second, they need find items quickly and select

them into temporary order list (just like shopping

trolley). Instead of writing the name of item, they

choose item from screen with a touch pen. Besides,

modification of order should be easy. Third, they

want to make sure items are served or not. So when

one item is put on table, it should be checked

simultaneously. Finally, it should be compatible to

conventional order process, and need little time to

learn the new system. If training time is too long,

neither managers nor waiters will think iMenu is a

successful application.

TOWARD PERVASIVE COMPUTING IN RESTAURANT

313

Customers need better services just at tableside.

By using iMenu, customers should get response

more quickly than before. It’s the best that ordering,

changing items, adding beverages, paying for foods,

everything can be done at tableside. With card

readers connected to mobile devices, waiters could

help customers finish payment transaction

effortlessly, which means customers can make

payment anywhere in the restaurant.

3.2 System Architecture

To satisfy all above requirements is a challenge.

However, as a pervasive computing application in

restaurants, iMenu does improve greatly the

experiences of both restaurant staff and customers.

The first design issue we concerned is to integrate it

into existing system. Since most medium to large

size restaurants use POS system nowadays, iMenu

should be able to access previous database. So,

multi-tiered web-based architecture becomes our

choice. Without touch any source code of previous

system, we integrate iMenu into the POS system

with an easy manner. System architecture of iMenu

is shown in figure 1. The hardware needed for this

system is a web server, wireless network, and some

mobile devices such as palm-top PDAs and

WebPads. The web server is a link between mobile

devices and existed database, which allows people to

manipulate the database wirelessly. Moreover, with

web services technology, it is also a bridge to

exchange data with other enterprises’ databases

located in difference places.

Figure 1: System architecture of iMenu

3.3 Interface

The second issue we concerned to design iMenu is

user interface. Input in small screen is not an easy

work, so minimizing interaction between people and

devices becomes our design emphasis. Nowadays

many PDAs integrate handwriting recognition,

which is very helpful when input characters.

However, to input a word is not enough at all, our

aim is in fact how to record quickly what customers

order. So, we use help code. With help code, waiters

can find “Live Lobster” by only input two letters “l”

and “l”. The beauty of help code is that it is

irrespective to what language people use, no matter

it is Germany or Greek. It could be just the initial

letter of each word’s pronunciation in their native

language, which is a convenient way to input and

remember. Other alternative methods are also

adopted, if waiters are familiar with restaurants

existing menu code system, they can use Arabic

numerals code. Or, the way that may cost more time,

they just input full item name and search. When

waiters find the item that customer orders, they tap

the item name. Then it is selected into a temporary

list. After confirm items with customers, waiters just

need click one button, everybody who needs the

ordering information can get it.

Figure 2: User interface of iMenu on PDAs

Figure 2 shows the interface of iMenu on PDAs.

We use Compaq iPAQ pocket PC as the mobile

client. In fact, any device with an Internet browser

can be used as a client. This is a clear advantage of a

web-based system. Figure 2(a) is the interface for

waiters to find and input items. When waiters

perform the search task, the search result list shows

in the screen; in other time the order list shows in the

screen. Figure 2(b) is a map of table status. We use

different colors to distinguish status. Blue means the

table is not occupied by anyone; yellow means the

table is booked by somebody; and pink means

currently there are customers occupying the table.

When blue button is stroked, an empty temporary

list is open for ordering (from point of view of

business, a new bill generated). While the red is

stroked, waiters can recall current information of

ordering of that table. Detail reservation information

can be found if the yellow is stroked.

ICETE 2004 - GLOBAL COMMUNICATION INFORMATION SYSTEMS AND SERVICES

314

We not only developed iMenu’s interface on

PDAs but also on WebPads. The high quality

display of WebPads expands the use of iMenu from

restaurant staff to customers. 10.4” TFT, touch-

screen with 800x600 resolution (SVGA) gives users

new ordering experience just like their own PC. If

users know how to surf on the Internet by computers,

they are already able to use iMenu. An exciting

merit is that we can embed multimedia features into

this system, since WebPad’s computing capability is

relatively better than small PDA. When customers

choose an item, the recipe, story, picture and even

movie about this item is shown. So customers may

enjoy ordering meals by themselves. Now,

interestingly, Waiser’s calm technology (Weiser,

1993) is not really “calm”, people can appreciate

movie and music before their dinner coming.

Shortcoming of WebPad is that it has heavy weight

to some extent. Although today mobile devices

become lighter and lighter, 1.7 kg is still too heavy

for waiters to carry for more than 2 hours. The So,

mainly it is used on table for customer self-service.

Figure 3 shows the WebPad we use in our system

and the interface of iMenu on WebPads. Since the

screen is large enough, we can display item

introduction and work zone in one screen.

3.4 Analysis

There are about 70 database tables used in our

system. Stored procedures and triggers are used to

fulfil some tasks. It could help to lighten the load of

web server.

Table 1: Database tables used in the system

Table Name Description

shiftdef Definition of restaurant staff shifts

auth_login Staff information, control of login

auth_group Staff group definition

pos_pda

Information of PDA, for example, name, address and

region

pos_tblsta Information of table

pos_reg

Information of region, one work region may have

several PDA

pos_tblav Information of table status

pos_pluid Menu category

pos_plu Menu

pos_dish_add Special cuisine requirement of guests

pos_dish_none Record of run-out items

pos_tempsession Template ordering list

During the development period, we find multi-

tiered architecture design has many benefits. Our

system needs the least investment on hardware and

maintenance. Before deploy the system, restaurant

owners would like to purchase the hardware as

cheap as possible. iMenu only needs thin client,

which is good for both technical and commercial

consideration. Any device with a browser can be

used as client. For instance, some kinds of PDAs

that have monochrome screen, which is inexpensive,

can be used as client devices. After deploy the

system, restaurant owners would like save cost on

maintenance as much as possible. Our system need

almost no maintenance work on client side and the

central web server is easy to maintain. We find some

other systems store an inventory database and client

application in mobile devices, which make it very

difficult to upgrade, because every client-side

software should be re-install, and every client

database should be updated. Besides, client-side

applications increase the load of mobile devices that

require higher performance equipments, which have

higher prices.

To system developers, this web-based

architecture is also helpful. For example, it is simple

to obtain pleasant user interface by web page

designing. And it has good scalability because it is

composed of several clear layers such as data access

layer, business logic layer, business presentation

layer, and user interface layer. Another benefit is

that programmers can debug the source code on

computers, and then check the compatibility

between the web page and mobile device’s browser.

It is much easier than to debug the client program on

mobile device simulator first and on real mobile

device later.

Generally speaking, compared with other

client/server pervasive computing applications, our

mobile ordering system has many advantages as

follows.

 It is simple to manage, and therefore more

reliable, because it doesn’t need different

configurations and installations on different

client mobile devices;

 It is more secure, in that all the access should

pass the central authentication; while the

information stored on different mobile devices is

at risk, in that back-up and archiving procedures

are less stringent than on servers or mainframes;

 It supports universal access to applications and a

common user interface, decreasing training costs

and minimizing confusion to system use;

 It costs less to upgrade and change, in that no

applications have to be installed one device at a

time; and

 It accelerates application development, because

business rules can be focused on server side

programming.

Performance is concerned when we develop

iMenu. If the client device is strong enough, client

TOWARD PERVASIVE COMPUTING IN RESTAURANT

315

applications running on client device can almost

response real-time. So we tested the response time of

iMenu to see the performance of web-base

application. We find the response time of every

request is no more than 0.7 second. And the load of

web server is not heavy for intranet use, because the

number of connections at one time is small. Besides,

we avoid using complicated script in web page when

programming, for alleviating load of both server and

clients, and providing better compatibility to client

browser.

3.5 Evaluation

We have deployed iMenu in two restaurant

enterprises. Then we conducted an evaluation study

focusing on the staff performance of ordering and

consumer experience. The first quarter after

deployment both of these two enterprises make more

than eleven percent profit than last year. We found

that by using this system the time staff spent in

service per customer reduced significantly.

According to the observation of 151 waiters in the

first hotel group in 1 month, the average ordering

cycle time per table is shortened from 12 minutes to

10 minutes; the waiting time after order is reduced

from 7 minutes to 5 minutes. The same results are

obtained in the chain enterprise group’s restaurants.

Not only time is saved, but working efficiency is

also improved. With iMenu, staff in restaurants can

cooperate more easily and smoothly. All the

ordering information is transferred wirelessly, with

least effort staff can work precisely. Moreover, staff

can communicate with short messages on mobile

device, but all for free, not like mobile phone text

message service that charges according to the

message sent. Therefore, they work more efficiently

than before.

On the side of consumers, it also greatly

improves the ordering experience. With iMenu,

people can get better and faster services, which is

the most desirable in restaurants. Moreover, people

may try to order dinner easily by themselves on

WebPad. Many young people would like this mode

of service.

The system is compared with other wireless

ordering systems. System 1 and system 2 are

wireless products of current world leading

companies in industry. System 1 is based on Pocket

PC, with 801.11b and integrating with the

company’s restaurant information system. System 2

uses Infrared as communication channel, with local

database and client application installed in mobile

device. We choose each system’s latest version.

Compared with these current commercial ordering

systems, iMenu has lots of advantages. The most

strength of iMenu is its scalability and client

operating system independence, which gives

enterprise much freedom to deploy and expand the

system. It also has better multimedia and integration

features than others.

Table 2: Comparison with other systems

3.6 Future Development

With pervasive computing, using information

technology will progressively feel more like using

the everyday objects than using personal computers

(Fano, 2002). Nowadays people may notice that the

restaurant using mobile device instead of traditional

menu, but later when the way of accessing and using

information is totally changed, people will not think

iMenu is a fangle of information technology.

Although our research successfully brings people

new ideas of restaurant ordering, we still have lots of

challenges in future development.

First of all, pervasive computing implicates that

devices are aware of their surroundings and peers,

and to be capable of effectively providing services

to, and using services from, peers (Cheng, 2002).

While our application still has low rating in device

discovery and service discovery because it uses

predefined network address and predefined

application specific access. In this mode, it is not

difficult to manage the limited number of mobile

devices. However, when there are a huge number of

devices, say, “pervasive”, only the configuration

work of IP address, subnet masks, gateways and so

on would become a very cumbersome job. We need

automated techniques to dynamically reconfigure the

network when required (Saha, 2003). We noticed

Features System 1 System 2 iMenu

Infrastructure

Wireless

Client/Server

Wireless

Client/Server

Wireless Multi-

tiered web-based

Wireless networking IEEE 801.11b Infrared IEEE 801.11b

Mobile devices PDA PDA

PDA and

WebPad

Client operating

system

Windows CE Palm OS OS independent

Multimedia features No No Yes

Pervasive database Integrated DB Local DB Integrated DB

User authentication Password Password Password

Integrated Reservation

System

No No Yes

Integrated Credit Card

System

No No

Mobile card

reader

Integrated Enterprise

Information System

Client access No

Web-based

access

ICETE 2004 - GLOBAL COMMUNICATION INFORMATION SYSTEMS AND SERVICES

316

that pervasive computing retail system applications

may offer benefits for both retail consumers and

suppliers (Kourouthanassis, 2003). So next step we

plan to develop a radio frequency identification

(RFID) system to detect the distribution of dishes in

restaurant, which will help to sense the service event

and capture related data without human interaction

with system. Since dishes with RFID tag of

restaurant are reusable, it has no cost problem like

retail system.

Secondly, security is another important and

complicated challenge related to our work. In

addition to contending with the usual Internet

security threats in online applications, wireless

devices introduce new hazards specific to their

mobility and communication medium (Ghosh,

2001). Now we mostly rely on built-in wireless

hardware security features and Secure HTTP

(HTTPS) implemented over Secure Sockets Layer

(SSL). However, to provide a secure pervasive

computing environment is not an easy job. Some

conceptual models of pervasive computing

environment are introduced (Kagal, 2001). But we

still face practical security problems such as insider

attacks, man-in-middle attacks, viruses attacks,

Distributed Denial-of-Service (DDoS) attacks

(Xiang, 2004) and so on. DDoS is possible in

wireless realm, and is as dangerous as in wired

Internet.

Security and privacy are intimately linked

together. From the view of technology, privacy can

be protected by secure communication channels and

data storage. Beside the technical issues, there are

also social issues on security and privacy. By

pervasive computing, more customer information

can be easily collected than before. So how to use

the personal data gathered by pervasive system still

need cultural or legislation solution (Beresford,

2003).

4 CONCLUSION

In this paper, iMenu, a web-based wireless ordering

system is presented. As we have seen, it has lots of

advantages such as offering customers fast and

accurate services, reducing restaurant staff working

errors, providing enterprise owners more profits and

so on. However, this is only the first step toward

realizing pervasive computing at tableside. Some

problems such as dynamic reconfigurable adaptation

of mobile devices, security and privacy issues,

scalabilities, integration to existed enterprise

information systems and so on remain open for

research.

REFERENCES

Beresford, A., R. and Stajano, F., 2003. Location Privacy

in Pervasive Computing, In IEEE Pervasive

Computing, 2(1).

Catarci, T., 1999. Web-Based Information Access, In 4th

IECIS Int’l Conf. on Cooperative Information Systems.

Cheng L. and Marsic, I., 2002 Piecewise Network

Awareness Service for Wireless/Mobile Pervasive

Computing, In Mobile Networks and Applications,

7(4).

Fano, A. and Gershman, A, 2002.The Future of Business

Services in the Age of Ubiquitous Computing, In

Comm. of the ACM, 45(12).

Ghosh, A., K. and Swaminatha, T., M., 2001. Software

Security and Privacy Risks in Mobile E-commerce, In

Comm. of the ACM, 44(2).

Kagal, L., Finin T. and Joshi, A., 2001. Trust-Based

Security in Pervasive Computing Environments, In

IEEE Computer, 34(12).

Kourouthanassis P. and Roussos, G., 2003. Developing

Consumer-Friendly Pervasive Retail Systems, In IEEE

Pervasive Computing, 2(2).

Noble, B., 2000. System Support for Mobile, Adaptive

Applications, In IEEE Personal Comm., 7(1).

Saha, D. and Mukherjee, A., 2003. Pervasive Computing:

A Paradigm for the 21st Century, In IEEE Computer,

36(3).

Satyanarayanan, M., 2001. Pervasive Computing: Vision

and Challenges, In IEEE Personal Communications,

8(4).

Stanford, V., 2003. Pervasive Computing Puts Food on the

Table, In IEEE Pervasive Computing, 2(1).

Weiser, M., 1993. Some Computer Science Issues in

Ubiquitous Computing, In Comm. of the ACM, 36(7).

Xiang, Y., Zhou, W., and Chowdhury, M., 2003. GHIRS:

Integration of Hotel Management Systems by Web

Services, In 2

Int’l Workshop on Grid and

Cooperative Computing (GCC2003), LNCS 3033.

Xiang, Y., Zhou, W., and Chowdhury, M., 2004. A Survey

of Active and Passive Defence Mechanisms against

DDoS Attacks, Technical Report, TR C04/02, School

of Information Technology, Deakin University,

Australia.

TOWARD PERVASIVE COMPUTING IN RESTAURANT

317