Public Perception of UAMs Study
Study Duration: Sept. 2020- Jan. 2021
Publish Date: October 2022, HFES Conference Paper
I am the first author for my paper on the Investigation of Perceived External UAM Design Features as a New Transportation Method, that was submitted to the Human Factors and Ergonomics Society annual meeting in 2022 where it won the award for Best Student Paper for the Product Design Technical Group. This was my first time doing UX research as part of the MUX Lab at the University of Michigan-Dearborn, and I thoroughly enjoyed the process!
Urban Air Mobility vehicles (UAMs) are a new method of proposed transportation in metropolitan areas to solve increasing traffic congestion issues due to rising populations in these areas. UAMs are flying vehicles that use electric vertical take-off and landing (eVTOL) technology and directional rotors to allow the vehicle to fly over traffic in congested areas. There have been many proposed use cases for UAMs, with the most common ones being for emergency medical services, parcel delivery, and even human transport. Below is a video showing some of those common use cases for UAMs made by some of my research teammates!
The Problem
UAMs are currently almost entirely conceptual, with the first few models being built in a few places around the world. Because of this, there is no standard image of what a UAM looks like for the average person. We wanted to compile existing UAM models to test which external design factors evoke different emotions in potential users.
Kansei Engineering
Kansei Engineering which is more commonly known as the marriage of engineering and emotion, founded by Dr. Nagamachi in the 1970s in Japan to reflect a measure of a system's feelings on to product design. Kansei engineering methods improve or develop solutions using the user's psychological feelings and needs and translating them into aspects of the product or service's design. We heavily relied on Kansei Engineering methods in our study, as we generated a list of possible semantic pairs of words that could accurately describe how one might feel after viewing a specific external feature of a UAM.
Study Objectives
1. Investigate common perceptions of external UAM design features and user preferences through Kansei Engineering.
2. Find common descriptor terms for UAM's external design.
3. Illustrate the descriptor terms in a 2-dimensional conceptual map.
4. Help the standardization of UAM designs.
We began by doing a literature review on existing literature of Urban Air Mobility vehicles (UAMs). We also collected hundreds of existing concepts and models, compiling and classifying them into various groups based on their external design features such as having wings or not, the position of the rotors, and if the rims were protected and exposed.
A
C
B
D
After compiling and classifying, we decided to go with these four models for our study. The models were chosen subjectively, but they included a different combination of the three classifying factors: rotors on the wings or the body, rotor position being on top or on the bottom of the body, and the shape of the rotor being protected or exposed. After picking the four designs, we sought out to create our measuring metric. The designs chosen did not take into account engineering or technical feasibility of flight, we only focused on external design features.
This is the list of 30 common semantic pairs we generated that could potentially be descriptor terms for different UAM external design features. The words selected for the list were also selected subjectively, using logic for what we thought could describe external design features of a vehicle.
We then created a google form with a static image of each model and asked participants to rate each model using the descriptor terms on a scale of 1-10. An example of this would be having a score of "1" being a user felt it was the safest the design could be, and a score of "10" being a user felt the design appeared to be the most unsafe.
We then created a google form with a static image of each model and asked participants to rate each model using the descriptor terms on a scale of 1-10. An example of this would be having a score of "1" being a user felt it was the safest the design could be, and a score of "10" being a user felt the design appeared to be the most unsafe.
​
We had 42 respondents, 21 male and 21 female, with 50% of the population being 18-30 years old, 25% of the population being 31-40 years old, and the last 25% of the population between 41-51, and 51+ years old.
Study Results
We conducted a factor analysis to obtain latent factors among the 30 semantic pairs, as well as determine how many factors would be necessary to describe the external design features of UAM designs. The ratings for the pairs were submitted to a principal component analysis with an orthogonal Varimax rotation. Through the analysis of the scree plot and the cumulative eigenvalues, five factors were identified as latent factors, explaining 70% of the total variance of the external UAM designs. Normally, we would have chosen the pairs with the highest factor loading variables, but after internal discussion, we chose what we thought were the simplest and most intuitive terms to describe the external design.
​
The Five Latent Factors are:
1. Safe/Unsafe: refers to how participants rated the different models on how safe the external design appeared.
2. Comfortable/Uncomfortable: refers to how comfortable the design appears to be.
3.Novel/Usual: refers to how new the design appears and if it gave users a sense of originality or not.
4. Simple/Complicated: refers to overall level of perceived complexity in the external design.
5. Basic/Advanced: refers to how futuristic, and advanced the design appears to be.
​
The full factor analysis results as well as the selected latent factors are listed in bold in the table below.
To meet our third goal of illustrating the descriptor terms on a 2D conceptual map, we conducted a multidimensional preference analysis (MDPREF) based on our principal component analysis. This allows us to layout the sample models in the same space as user's ratings of them based on the five latent factors. The MDPREF is a graph of ideal points, with the five latent factors being represented as vectors. A general preference vector for overall "like" and "dislike" of a product was also included to establish a relationship between the design samples and the descriptor terms with overall general preference.
Based on this MDPREF, we can gather a few findings about each design and how their external designs made users feel.
​
Model A was seen as the most basic and usual, with many users commenting that it is similar to drones they have seen in the market today. Model A was also perceived to be less safe than Model D and Model B, suggesting that along with Model C, unprotected rotors makes users feel like the design could be less safe than the protected rotors.
​
Model B further supports the idea that protected rotors are perceived to be safer and generally more preferred, and it also appeared to be more comfortable than Model A and Model C.
​
Model C was the most novel, unsafe, complicated, uncomfortable, advanced, and overall most disliked design. It was also consistent with the findings that unprotected rotors are not preferred. The placement of Model C on the MDPREF suggests the design is just too novel and advanced, and users would like to see some degree of familiarity with a vehicle such as a UAM.
​
Model D was the most liked design, and it was perceived to be the safest, comfortable, and simplest design. Placing closest to the general preference vector, this design suggests that users may like having the rotors being placed above the body so there is no potential harm in running into them, a large cabin for comfort, and continues to support the finding that protected rotors were generally preferred.
Overall, safety seems to be most important factor in determining if a user would actually utilize UAM technology, which is consistent with the finding of The Booz Allen Team, that 93% of focus group participants would be willing to us a UAM if their safety was guaranteed.
Conclusion
This study is foundational work in trying to standardize a model for UAMs. Through our findings, we recommend UAM designers to prioritize perceived safety from external design features, adding things like protected rotors, placing the rotors above the body, and having a large enough cabin for the passengers to feel like they would be comfortable in the vehicle.
​
Some of the limitations for this study include having a small sample size, only presenting static images with different backgrounds and only one angle of the models, and also not taking into account any engineering or technical feasibility of the models.
​
For future research, we are looking to conduct the study in a VR environment with 3D models that users could interact with and view from any angle. We have already conducted this testing and will look to publish the results soon!
Conference
This paper was presented at the Human Factors and Ergonomics Society 2022 Annual Meeting in Atlanta, GA. This paper won the award of "Best Student Paper" in the Product Design Technical Group and a copy of the paper can be found here.
