THE GROCER PROJECT - TOWARDS TO A REAL AUGMENTED-COMMERCE ENVIRONMENT
Pedro Teixeira[1] Enabler Informática, Sa, Portugal,
Jorge Santos[2] Enabler Informática, Sa Portugal,
Orlando Belo[3] Universidade Do Minho, Portugal

Abstract

GROCER is a pioneering project that enables the retailer to create a long-term two-way relationship with the customer, combining both physical and virtual shopping experience. It allows a high level two-way interaction with the customer that could drive customer behaviour changes, an iterative context over time, an ongoing benefit to both parties, and a unique approach to the individual customer. In this paper, we present the GROCER's system, addressing its context, primary goals, and the most important aspects of its functional architecture and software agents.

Keywords:  Augmented-Commerce, Retail Systems, Agent-Based Computing, Ubiquitous Computing, and JAVA Technology.

1.       Introduction

Retail profitability is now tightly connected to customer insight. We live in an information era where the customer is king. We can talk about �The Customer�s Victory�. The leading retailers need to build a powerful customer insight to focus their merchandising, marketing and customer service offerings into a powerful and integrated brand offering. This is critical to leverage what is becoming clearly their most critical asset to leverage � customers. Retailers that successfully integrate the different channels into a singular and unique face to the customer will be able to build two-way relationships with them that create strong switching costs, thereby establishing a sustainable competitive advantage. Today�s customer driven economy requires revolutionary methods of differentiating the business in a crowded market. Once having this customer insight, is critical for the retailer to be able to deliver it in a smooth way to the customer preferred touch point � whether that�s web, email, catalogue, in-store, phone, etc. � or any combination of these.

In the last few years, we have been assisting to a massive adoption of mobile technologies by people. From laptop computers to compact palm-sized organizers, and, naturally, cellular phones, people use them to do a lot of things, anywhere, any time, and when they want. Today, it is not difficult to find someone using a WAP device trying to establish a remote Internet connection, reserve a cinema ticket for a movie in a specific room, consult a specific bank account system, or simple searching for information about goods and services. In a very near future all these activities will be done regularly by anyone that owns a mobile device, anywhere in the world. Many experts have the opinion that mobile devices are the future of personal computing. However, grocery stores have not experimented or explored yet this kind of technological wave in their commercial activities, promoting and selling goods and services. Some of traditional grocers have opened their own online stores in the Internet. With this event, new forms of shopping emerged. Customers have thus the possibility to order the products they need without going out, and, when implemented, they have access to some advising services. A new frontier was open when organizations tried to implement new forms of shopping over the Internet, supported by software agents. Some of these organizations developed serious efforts to put these entities doing business with customers in their sites. The introduction of agent-based technology inside commerce environments is not new. Since its first real application in information retrieving and filtering scenarios, the use of agents have been increased in other real applications.

Primarily, shopping agents [7] were specially conceived to act as a privileged form of interface, helping customers searching and comparing products, or promoting goods and services. BargainFinder [6], Shopper's Eye [3], or Magnet [2] were some of the projects developed under such efforts, which contributed to bring a new dimension to the shopping experience. With no doubt, agents brought great advantages to organizations. They improved the quality and availability of their attendance services, being available 24 hours a day, running continuously and doing things rationally and autonomously, ensuring and supporting a large variety of services � information gathering, searching products, pricing comparison, negotiating shopping terms, evaluating offerings, or receiving payments. A new generation of mediators emerged. Electronic commerce reinforced its position in the virtual world. However, online grocery stores have not enjoyed as much success as expected. There are many reasons that could justify such little success. Provably, when people use exclusively this new way of shopping, they do not have anymore the pleasure of going to a grocery store enjoying the �experience� of a shopping afternoon. Online stores have naturally their own space and are one of the most versatile ways of shopping that we have today. But we must think in other alternative ways, trying to conciliate innovative and sophisticated means with the pleasure of shopping. Some of the difficulties occurred in the past, provably were attenuated by the strong emergence of mobile and ubiquitous technologies. Today, people do not have anymore to be in front of a computer to access to a shop online. Rather, they could use their mobile phones, or any other kind of mobile device, when they want, and potentially in any place, to put their orders or simply get a price for a product. It is very usual people remember what they need to do or to buy, when they move around seeing things or interacting with other people. Using a mobile phone they could do what they need at the moment they thought. This enlarged enormously the potential of electronic commerce.

2.       The GROCER Project

2.1.         General Overview

The fundamental focus and essence of the GROCER[4] � Grocery Store Commerce Electronic Resource [1] � project is exactly to help on building this insight and delivering it on the channel that generates the biggest revenues � �The Store�. Grocer is a pioneering project that enables the retailer to create a long-term two-way relationship with the customer. GROCER combines physical with virtual shopping experience, bringing a new way of shopping to conventional retail environments. The system will identify the moment when the customer mind context is more open to an opportunity to buy. The final objective is to create the next frontier customer experience. GROCER allows a high level two-way interaction with the customer that could drive customer behaviour changes, an iterative context over time, an ongoing benefit to both parties and a unique approach to the individual customer. Creating an intimate and unique non intrusive relationship, taking care and paying attention to customer needs remembering and responding to them, will allow with time the creation of a key and fundamental aspect for the retailer: �The Customer Trust�. To achieve these objectives, the GROCER system will act at the different operational aspects of the store (Figure 1). Its functionalities focus on improving all the store operations involved in the customer experience, from the more basic ones (on shelf availability, price accuracy, queue busting, etc.) to the ones related with sales staff support ending on the more advanced (one-to-one voice interaction over IP, retailtainment, myShoppingAssistant, etc.).

Figure 1. Benefits, formats and services of GROCER.

GROCER will allow the retailer to build an evolutive strategy that will protect his investment.  Its highly flexible and scalable architecture based on components, will allow the retailer to start focusing for example in the aspects related with operations, moving to the ones related with store staff service to customers, ending on the advanced customer interaction.

2.2.         Design and Development Aspects

The project initiative is a pioneering experience in the field of augmented commerce. Its main objective was to provide the basis for the convergence of electronic and physical commerce services and infrastructures into a single system. A complete system inspired on this new model of commerce was designed and developed, providing the basis to combine physical with virtual shopping activities. The system provides means and mechanisms that customers can use through their mobile devices, in order to receive customized services or having support and advice during their shopping activities. Such characteristics will make shopping activities more easily enjoyed and afforded. Tasks such as parking, advertising, recipe suggestion, merchant brokering, or product location will be possible to do inside a Grocer environment, using adequate interfaces.  Some of the aspects that we considered during the design and development of GROCER were to:

Develop and implement virtual retail markets, configured for specific merchants, where characteristics such as commodity, efficiency, openness, flexibility, security and direct negotiation are basic issues on the support of customers on shopping activities.

Provide easy and flexible access to virtual retail markets through mobile devices and platforms (e.g. personal digital assistants, mobile phones, lap top computers, etc.) as well as through wireless web platforms, or even through conventional user interfaces, making possible the access to a shopping environment virtually from anywhere improving welfare and customer's attendance on shopping activities.

Develop a new set of shopping tools (retail agents) that help consumers make buying decisions within a specific merchant's virtual market, comparing product offerings, retrieving merchant and product information to help to decide what to buy, helping in the evaluation of product and services alternatives based on consumer's criteria, system interaction or their shopping history, or simple receiving support about product location or about the best path to reach the goods they want to buy or see.

Design and implement a back-end retail services system with the abilities to receive, in a parallel virtual environment, other complementary agents oriented to support related shopping activities, namely recipe suggestion, fashion advising, dietetic recommendations, entertainment activities, or even to get help in case of some emergency situation.

Research a implement new models for customer profiling in the area of electronic commerce, developing a specific class of agents with the ability to learn user's preferences during shopping activities and build a profile without redundant or unnecessary efforts, avoiding users to fulfil large and complex questionnaires.

2.3.         The System Architecture

The functional architecture of GROCER (Figure 2) integrates six distinct platforms: 1) user interface, which ensures support for user-system interaction; 2) application server, that provides the means to access data and business rules, and controls user sessions; 3) agents' environment, a distributed platform that allows to receive the system�s agents, ensuring their communication and data exchange processes; 4) business and data access, a data-oriented platform specially conceived to store and manage system�s data and business rules; 5) merchandise and location services, which includes all the services relating tracking and positioning of customers inside stores, and query processing about product location or stores� plant; and 6) mobile user support, that is responsible to send messages to mobile services providers and ensure communication to mobile devices equipped with Bluetooth. Together, these modules ensure the functionalities needed to guarantee a real augmented-commerce environment.

Figure 2.  The GROCER�s Functional Architecture.

2.4.         The Agents

Software agent technology [5] [8] was used in order to automated, improve, and reduce human influence in the most time-consuming stages in shopping processes [4]. The use of software agents - seen as artificial workers - inside typical retail scenarios helps to optimise buying experiences and revolutionary the commerce, as we know today. Grocer's agents act basically as mediators, having the ability to act opportunistically and, when necessary, cooperate with each other. The main activities performed by agents inside a Grocer environment are: authenticating users and giving access to resources and customer oriented services, working on customer assistance, advising customers about the best way to find what they want, indicating the best path in the store to go shortly to a specific product location, launching season promotions, suggesting recipes according to a set of prerequisites, establishing shopping profiles or simply manage car parking places. Grocer�s agents interact with each other using a specific communication language that ensures communication and information sharing among them. This language was designed based on the FIPA (www.fipa.org) Agent Communication Language and integrates only a subset of the communication primitives that FIPA proposes. This subset was selected based on the services that GROCER provides.

In Figure 3 one can observe the base agent communication class that implements message transference and the basic structure of a message. Grocer�s agents are autonomous. They have the ability to control and management their own resources, promoting process cooperation with other agents when necessary.

2.5.         Implementation

GROCER was implemented entirely in Java, using SOAP, a protocol specification that defines a uniform way of passing XML-encoded data, as support for the interconnection services between the agent-based-platform and the other Grocer's modules. The agent-based platform, in particular the implementation of the distributed environment that support agent communication and data exchanging services was developed based on JINI technology, provided by Sun Microsystems. JINI is a set of Application Programming Interfaces and network protocols that enable the creation of distributed systems as federations of services/objects.

 Figure 3.  The base agent communication class.

3.       Conclusions

This paper presented the GROCER project. It combines physical with virtual shopping experience, bringing a new way of shopping to conventional retail environments. Additionally, GROCER also brings to real world retail scenarios an easy-to-use augmented-commerce platform, combining mobility, efficiency, shopping commodity, and service assistance in shopping platforms.

4.       References

Belo, O., Ribeiro, A., Santos, J., An Augmented-Commerce Environment for Real World Retail Applications. In Proceedings of the 7th WSEAS International Multiconference on CIRCUITS, SYSTEMS, COMMUNICATIONS and COMPUTERS, Corfu, Greece, July 7-10, 2003.

Collins, A., Tsvetovat, M., Mobasher, B., Gini, M., Magnet: A multi-agent contracting system for plan execution. In Proceedings of Workshop on Artificial Intelligence and Manufacturing: State of the Art and State of Practice, pages pp 63--68. AAAI Press, August 1998.

Fano, A., "SHOPPER'S EYE: Using Location-based Filtering for a Shopping Agent in the Physical World, In Proceedings of the Second International Conference on Autonomous Agents, 9-13 May 1998, pp. 416-421.

Guttman, R., Maes,P., Agent-mediated integrative negotiation for retail electronic commerce. In Pablo Noriega and Charles Sierra, editors, Proceedings of the 1st International Workshop on Agent Mediated Electronic Commerce (AMET-98), volume 1571 of LNAI, pages 70--90, Berlin, May 10--10 1999. Springer.

Jennings,N., (2001) "An agent-based approach for building complex software systems" Comms. of the ACM, 44 (4) 35-41.

Krulwich, B., The BargainFinder Agent: Comparison price shopping on the Internet. In Williams, J., ed., Bpts and Other Internet Beasties. SAMS.NET, 1996.

Maes, P., Guttman, R. H., and Moukas, A. G. (1999). Agents that buy and sell. Communications of the ACM, 42(3):81--91.

Wooldridge, M. e P. Ciancarini. �Agent-Oriented Software Engineering: The State of the Art�. In Paolo Ciancarini e Michael Wooldridge (editors), �Agent-Oriented Software Engineering�. Springer-Verlag Lecture Notes in AI Volume 1957. Janeiro, 2001.