The Turkish Ice Cream Robot: Examining Playful Deception in Social Human-Robot Interactions

Hyeonseong Kim, Roy El-Helou, Seungbeen Lee, Sungjoon Choi, Matthew PanarXiv 2025 · DOI · arXiv

Abstract

In human-robot interaction (HRI), deception is traditionally viewed as a barrier to trust and transparency. This research challenges that paradigm by exploring "playful deception"—specifically the teasing routine of Turkish Ice Cream (TIC) vendors. We developed a robotic system capable of performing evasive maneuvers during a handover task to evaluate the psychological impact on users. A study with 91 participants revealed that while deceptive behaviors significantly boost engagement, pleasure, and perceived robot competence, they simultaneously reduce perceived safety and moral trust. Notably, the entertainment value of the deception nearly doubled participants' willingness to pay for the product, highlighting a strategic trade-off for service robotics in entertainment contexts.

Introduction

Most robotic research focuses on predictability to foster user trust. However, human social life is enriched by benign violations of expectations, such as magic tricks or practical jokes. The "Turkish Ice Cream" vendor routine is a quintessential example of playful deception, where the vendor repeatedly snatches the cone away from the customer.

The robotic system mimicking the playful teasing of a Turkish Ice Cream vendor.

The authors argue that strictly avoiding deception may limit the potential of robots in hospitality and retail. By applying "Benign Violation Theory"—which suggests humor arises when a norm is violated in a safe context—this study examines whether a robot can be deceptive without being perceived as malicious.

System Implementation and Methods

The researchers designed a custom robotic platform to replicate the TIC experience. The system utilizes a UR5e industrial manipulator enhanced with a high-speed stepper motor to achieve the rapid twisting motions necessary for effective teasing.

System architecture including the control logic, hardware components, and safety mechanisms.

Hardware and Safety

To ensure user safety during deceptive interactions, the 55cm carbon-fiber rod and ice cream cone are attached via neodymium magnets. This "breakaway" design allows the components to detach safely if a user grabs the cone too forcefully or if a collision occurs. Hand tracking is performed via a dedicated camera to allow the robot to react to user movements.

Trick Policies

The robot was programmed with five distinct behaviors derived from observations of human vendors:

  1. Bouncing: Vertical movement of the cone.
  2. Twisting: Rapid rotation to evade a grasp.
  3. Large Arc: Wide sweeping motions.
  4. Continuous Arc: Fluid, circular evasive paths.
  5. Dancing: Rhythmic movements to engage the user.

Experimental Results

The study employed a mixed-design approach (N=91), comparing a "Straight Handover" (SH) control group with a "Deceptive Handover" (DH) group across varying interaction durations.

Enjoyment vs. Trust

The DH routine scored significantly higher in categories of Pleasure, Challenge, and Engagement. However, this came at a cost: the SH group rated the robot significantly higher in Reliability, Ethics, Transparency, and Perceived Safety. This confirms a fundamental trade-off: playful deception increases fun but erodes traditional trust metrics.

Competence Asymmetry

An unexpected finding was the "competence asymmetry." Participants rated the robot as more competent when it successfully tricked them, viewing the evasive maneuvers as a sign of high technical skill. Conversely, users' "self-competence" scores dropped, as they felt less capable when failing to catch the cone.

Economic Value

Despite the decreased efficiency of the deceptive interaction, participants in the "Long" duration DH group demonstrated a significantly higher Willingness to Pay (WTP). The average WTP for the ice cream was $9.87 USD in the deceptive group, compared to $4.59 USD in the straight handover group, suggesting that the "performance" itself adds substantial commercial value.

Discussion and Conclusion

This research demonstrates that benign deception can be a powerful tool for creating engaging and memorable robotic interactions. While it negatively impacts perceived safety and moral trust, the gains in enjoyment and perceived competence are significant for entertainment-oriented applications.

Future work will focus on making the robot more adaptive by integrating real-time sensing of user frustration or joy. While playful deception is effective in a retail "show" context, the authors caution that these findings are context-specific and would not translate to industrial or medical settings where predictability remains paramount.

Summary updated

Abstract

Playful deception, a common feature in human social interactions, remains underexplored in Human-Robot Interaction (HRI). Inspired by the Turkish Ice Cream (TIC) vendor routine, we investigate how bounded, culturally familiar forms of deception influence user trust, enjoyment, and engagement during robotic handovers. We design a robotic manipulator equipped with a custom end-effector and implement five TIC-inspired trick policies that deceptively delay the handover of an ice cream-shaped object. Through a mixed-design user study with 91 participants, we evaluate the effects of playful deception and interaction duration on user experience. Results reveal that TIC-inspired deception significantly enhances enjoyment and engagement, though reduces perceived safety and trust, suggesting a structured trade-off across the multi-dimensional aspects. Our findings demonstrate that playful deception can be a valuable design strategy for interactive robots in entertainment and engagement-focused contexts, while underscoring the importance of deliberate consideration of its complex trade-offs. You can find more information, including demonstration videos, on https://hyeonseong-kim98.github.io/turkish-ice-cream-robot/ .
BibTeX
@misc{kim2025turkish,
  author = {Hyeonseong Kim and Roy El-Helou and Seungbeen Lee and Sungjoon Choi and Matthew Pan},
  title = {{The Turkish Ice Cream Robot: Examining Playful Deception in Social Human-Robot Interactions}},
  year = {2025},
  eprint = {2509.21776},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.21776},
}
Figure