CHI 2026 · Barcelona, Spain

Understanding User Requirements for Creating Sensor-Powered Smart Car Cabins Through Retrofitting

Bofan Yu, BoRui Li, Tingyu Zhang, and Xing-Dong Yang

Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems

PDF DOI
Three-part illustration showing a user repositioning sensors inside a car, carrying a sensor setup between vehicles, and installing sensors in another vehicle
Figure 1. Retrofitting lets people select and position sensors for a particular cabin, reconfigure that setup over time, and carry familiar smart-cabin experiences between vehicles.

Abstract

In this paper, we explore a novel approach that leverages retrofitting to create sensor-powered smart car cabins. We propose that retrofitting offers a promising way to complement and extend the capabilities of built-in smart cabin sensors provided by car manufacturers. To understand how retrofitting solutions should be designed, we conducted a two-phase study. First, through semi-structured interviews with 18 participants, we examined challenges with built-in smart cabin sensors and identified opportunities where retrofitting could address these limitations. Second, through probe-based participatory design sessions with 15 participants, we identified user requirements and expectations for effective retrofit solutions. Based on our findings, we present a set of design recommendations to guide the future development of retrofit methods for smart car cabins.

Study design

From everyday limitations to situated co-design

We began with interviews to understand where factory-installed cabin sensors fall short, then moved into the car itself so participants could explore what a configurable retrofit system should support. This sequence connected broad concerns about customization, repair, upgrades, and portability to concrete installation decisions.

Timeline of a two-phase study with semi-structured interviews followed by in-car probe-based co-design
Figure 2. Phase 1 combined a short orientation with scenario-based interviews involving 18 participants. Phase 2 paired in-car familiarization with a discussion-card co-design activity involving 15 participants.

Design probes

Making sensor choices tangible

Physical mockups turned an abstract idea into something participants could hold, move, and debate. By trying placements from different seats, participants could reason about visibility, reach, sensing performance, privacy, driving safety, and how easily a setup could be removed or transferred.

Nine 3D-printed sensor mockups representing a camera, GPS, microphone, gesture, light, humidity, millimeter-wave radar, pressure, and temperature sensors
Figure 3. Nine 3D-printed sensor probes represented familiar sensing modalities. Velcro-backed mounts made it easy to compare locations without permanently altering the vehicle.
Participants placing sensor mockups on the dashboard, steering wheel area, roof, center console, rear seats, and other locations inside a car
Figure 4. Participants tested the probes throughout a real cabin, revealing how preferred locations depend on the sensor, the occupant, the vehicle layout, and the situation.