A Two-Phase Fuzzy Quality Function Deployment Framework for Customer-Driven Intelligent Design of Hotel Service Robots

Imagine walking into a hotel lobby and seeing a robot glide smoothly to your side, offering a warm welcome, guiding you to your room, and perhaps even fetching a bottle of water. For this interaction to feel natural and safe, the machine must do more than just follow a pre-programmed path; it must understand the world around it, recognize a person versus a wall, react politely to a child running by, and respect the privacy of everyone it meets. This is the challenge of designing intelligent service robots for hospitality. While engineers have long known how to build machines that move and lift, teaching them to navigate the unpredictable, social world of a hotel with the same grace as a human concierge remains a complex puzzle. The difficulty lies not just in the hardware, but in translating vague human desires—like "be safe," "be polite," or "adapt to the crowd"—into concrete engineering instructions that a computer can actually follow.

A team of researchers from Quanzhou University of Information Engineering in China has proposed a new way to solve this translation problem. They developed a structured method to take the subjective wishes of hotel guests and managers and turn them into a clear, prioritized list of technical capabilities. Instead of guessing which features matter most, the team used a two-step process involving a group of experts to map out exactly what a hotel robot needs to do before deciding which technologies to build. Their approach treats the design process as a bridge, connecting the "what" of customer needs to the "how" of engineering solutions, ensuring that the final machine is driven by human expectations rather than just technical convenience.

The researchers began by gathering the voices of those who would actually use or manage these robots. They brought together twenty-two people, including hotel managers, frequent guests who had stayed in hotels many times, robotics engineers, and specialists in artificial intelligence ethics. These participants were asked to rate the importance of various customer requirements, such as the robot's ability to move safely around people, its speed in responding to requests, and its capacity to handle data with privacy. The team then translated these linguistic ratings—words like "very important" or "fair"—into a mathematical format that could handle the natural uncertainty and vagueness of human opinion. This allowed them to calculate a precise score for how much each requirement mattered, even when different people had slightly different views.

The first major step in their framework was to convert these customer requirements into "intelligent design attributes." These are the specific cognitive and behavioral capabilities the robot must possess. For example, the desire for a robot that moves safely around guests translates into the attribute of "safe human-robot interaction." The wish for a robot that adapts to a busy lobby becomes "adaptive behavior." By running their data through this first phase, the team discovered a clear hierarchy of what matters most. The highest priority was "context awareness and perception," meaning the robot's ability to sense and understand its surroundings. This was followed closely by "safe human-robot interaction" and "operational reliability." Interestingly, while features like a friendly voice or a sleek appearance were important, they ranked lower than the fundamental need for the robot to be safe, reliable, and aware of its environment.

Once these priorities were established, the researchers moved to the second phase: determining which specific technologies would enable these capabilities. They asked the engineering experts to link the prioritized attributes to actual technical solutions. The result was a ranked list of "enabling technologies." The number one technology identified was "AI-based perception and control," which provides the robot with its core ability to see and think. The second most critical technology was "sensor fusion and autonomous navigation," which combines data from cameras and other sensors to help the robot move without crashing. The third was "machine learning for adaptive behavior," allowing the robot to improve its actions over time based on experience. The study showed that to build a successful hotel robot, developers must focus first on these foundational technologies that ensure safety and reliability, rather than rushing to add complex social features that might fail if the robot cannot navigate the room safely.

The researchers also paid close attention to the ethical side of artificial intelligence, a concern that is often overlooked in early design stages. They included a specific category for "ethical AI compliance," which covers transparency, data privacy, and fairness. Their analysis revealed a fascinating tension: the more a robot tries to understand its environment in detail (to be safe and helpful), the more data it collects, which can threaten guest privacy. The team found that this trade-off is a critical design constraint. They suggested that engineers must design systems that can process data locally on the robot itself, rather than sending it to a central server, to maintain both high awareness and high privacy. This insight ensures that ethical considerations are not an afterthought but are built into the robot's core architecture from the very beginning.

To ensure their findings were robust, the team tested how their rankings would change if the importance of certain customer requirements shifted slightly. They simulated scenarios where guests cared more about speed or managers cared more about maintenance. Even with these changes, the top three priorities—awareness, safety, and reliability—remained stable. This suggests that the need for a robot to be safe and reliable is a universal requirement for hotel service, regardless of specific cultural or regional differences. While the study was conducted with experts in Fujian Province, China, the stability of these results implies that the fundamental engineering needs for such robots are likely consistent across different hotel environments.

The study does not claim to have built a perfect robot or to have solved every problem in the field. The work presented is a methodological demonstration, a blueprint for how to think about the design process rather than a final product. The researchers explicitly noted that their results are based on expert opinions and simulations, not on testing a physical robot in a real hotel. They also acknowledged that their model simplifies some complex relationships between technologies, treating them as linear connections when, in reality, they might interact in more complicated ways. Furthermore, while they identified ethical AI as a key attribute, they did not have enough data to fully quantify how ethical performance should mathematically adjust the final rankings, leaving that as an area for future research.

Despite these limitations, the value of the work lies in its clarity and structure. It offers a traceable path from a guest's wish for a safe, polite experience to the specific sensors and software algorithms required to deliver it. By breaking the design process into two distinct phases, the framework prevents engineers from skipping steps or prioritizing flashy features over essential safety systems. It provides a common language for hotel managers, who know what guests want, and robotics engineers, who know what can be built, to work together effectively. The study concludes that for intelligent service robots to succeed in the real world, they must be grounded in a deep understanding of human needs, with safety and reliability serving as the non-negotiable foundation upon which all other features are built.