A Framework for Understanding the Experiences of Consumers with Humanoid Service Robots

Dr Robin Nunkoo

Professor

University of Mauritius

r.nunkoo@uom.ac.mu

Introduction

• Artificially intelligent (AI) service robots include various types of robotic devices that serve humans.
• Development and utilization of service robots has been gaining momentum during the last few years.
• Organizations in the service sector, such as hotels have also started utilizing AI technologies such as service robots in their service production and delivery processes. For example, Hilton Worldwide employs a robotic concierge named "Connie", who is responsible for personalizing guests' experience.
• Some hotels in Shanghai use AI robot servers that speak different languages to provide in-room services (Holley, 2019).

Applications of AI in Service Delivery (cont.)

• AI agents in services:

Smart devices

Self-service technologies

Chatbots

ChatGPT

Metaverse

AI Service robots

Social robots interaction

Humanoid service robots

3

The Problem

• Service robots can provide not only more consistent and timely service, but also a higher quality of service compared to human employees due to the advanced data storage capabilities empowered by AI technology, high processing speeds, and accurate personalization features (West et al., 2018).
• However, research has found several challenges in the use service robots in the service delivery process.
• One of the most prominent one is consumers’ acceptance of service robots across various context.
• The full potential of AI devices can only be achieved if users have positive attitudes toward and accept its use in the service delivery process (Belanche et al., 2020; Gursoy et al., 2019; Lu et al., 2019).

The Problem

• However, studies that examined customers’ intention to use AI devices during service delivery are mainly based on existing technology acceptance theories.
• For example, Sundar, Waddell, and Jung, (2016)’s study regarding AI robot adoption intention was based on the Technology Acceptance Model (TAM) developed by Davis, Bogozzi, and Warshaw in 1989.
• Even though traditional technology acceptance models, to a certain extent, explain the mechanism of users’ intention to use AI devices, these models were originally developed to investigate the adoption of non-intelligent technologies such as self-service technologies in service delivery (Im & Hancer, 2017; Lee & Cranage, 2018), whereas AI devices possess humanlike intelligence powered by AI technology.

The Problem

• AI devices such as service robots do not require users to learn how to operate them, which makes the ease-of-use construct included in all technology acceptance models as a core construct, irrelevant to the examination of customer's willingness to accept the use of AI devices during a service encounter (Lu et al., 2019).
• Furthermore, service are designed to interact with customers like regular frontline employees, and thus the construct of perceived usefulness included in previous technology acceptance models is also not likely to be relevant to predict customers’ willingness to accept the use of AI devices during service encounters (Lu et al., 2019).

The Problem

• Using AI devices to replace employees not only concerns task performance at the physical level, but also serves the social (i.e., AI used as automated social presence) and emotional meaning embedded in service interactions (van Doorn et al., 2017).

​

• However, none of existing acceptance model is suitable to analyze acceptance of a technology that characterizes the multi-faceted role of AI devices used in service encounters.

Our Project

12

Meet my dear friend Furhat

the social robot

Innovative Aspect of the Project

�Our Framework

• Based on a three-staged process of adoption/rejection based on Lazarus's cognition-motivation-emotion framework (1991a) and the major predictors of consumers’ willingness to integrate AI service robots into service delivery.

​

• Lazarus's framework suggests that during a decision-making process, individuals go through several stages of appraisals, including evaluating importance (primary appraisal), analyzing behavioral options (secondary appraisal), and creating emotions toward the stimulus that leads to behavioral intentions.

Our Framework

Applications of the AIDUA Framework

Application of the AIDUA Framework

• The Framework has been applied, fully or partly, in the US and China mostly.
• It has also been extended to include new variables such as trust and culture.