Designing AI-enabled solutions to support exercise motivation and active lifestyle: Case: Virtual Fitness Trainer
Valtanen, Pinja (2026)
2026
Tietojenkäsittelyopin maisteriohjelma - Master's Programme in Computer Science
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2026-05-29
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202605226182
https://urn.fi/URN:NBN:fi:tuni-202605226182
Tiivistelmä
Artificial intelligence (AI) has risen in popularity during the past decade with promising integration in the fields of physical exercise, nutrition, and health applications, often supported by persuasive design to steer users’ behaviour towards improved lifestyle choices. As a more sedentary lifestyle has become the norm for many after the COVID-19 pandemic and the increase in remote work, researching the persuasiveness of fitness applications with AI-enabled functionalities is needed.
This thesis aims to fill the gap between persuasive design for exercise applications and the integration of artificial intelligence, with the goal of investigating how an AI-enabled virtual trainer could support user motivation for exercise. Motivation is theorized using Self-Determination Theory’s three basic psychological needs: autonomy, competence, and relatedness. Persuasive design principles are presented to illustrate the importance of motivational aspects in designing a supportive AI-enabled fitness application, and the potential of AI in such a context.
To reach the goal of supporting user motivation through persuasive design in an AI-fitness trainer application, two research questions were formulated: 1) What are users’ initial expectations of an AI-enabled fitness trainer for supporting exercise motivation? and 2) How can an AI-enabled fitness trainer be designed to enhance exercise motivation and promote sustained physical activity? To explore the topic in more detail, users’ ethical concerns towards AI, as well as goals for exercise, were also researched. The qualitative study was conducted through semi-structured user interviews with six volunteering participants.
The results show that participants preferred functionalities such as the ability to track and predict their fitness development with statistics and gather achievements from completed workouts, aligning with the persuasive design principles of “self-monitoring” and “rewards.” Seeing the sources of the AI-enabled fitness trainer’s recommendations was also highly valued, relating to the principle of “credibility.”
Less preferred or unmotivating persuasive functionalities were “reminders” about daily exercise, as well as “social comparison,” especially when not perceived as lighthearted fun within a friend group. Participants were divided between those who preferred exercising alone and those who gained motivation from social factors such as peer pressure and shared activity challenges.
Most proposed AI-enabled fitness trainer functionalities had both positive and negative feedback from the participants, indicating that motivation is highly based on personal preference. Environmental context, current energy levels, and busyness could also hinder the motivation to exercise. From this, it can be inferred that the AI-enabled fitness trainer should adapt to varying situations and preferences, thus offering highly personalized support.
Based on the findings, collectively preferred functionalities were combined into a table as design guidelines for the AI-enabled fitness trainer. Additionally, UI wireframes to visualize examples of design decisions based on these guidelines were implemented using Figma. The UI designs highlight persuasive elements such as modifiable exercise programs, rewards, and positive feedback.
This thesis offers a new perspective on combining persuasive design principles with AI-enabled fitness trainers. By better understanding the connection between persuasiveness and motivation in AI-enabled trainers, users can be better supported in progressing toward their individual fitness goals—whether that involves walking up stairs without losing breath or running a marathon.
This thesis aims to fill the gap between persuasive design for exercise applications and the integration of artificial intelligence, with the goal of investigating how an AI-enabled virtual trainer could support user motivation for exercise. Motivation is theorized using Self-Determination Theory’s three basic psychological needs: autonomy, competence, and relatedness. Persuasive design principles are presented to illustrate the importance of motivational aspects in designing a supportive AI-enabled fitness application, and the potential of AI in such a context.
To reach the goal of supporting user motivation through persuasive design in an AI-fitness trainer application, two research questions were formulated: 1) What are users’ initial expectations of an AI-enabled fitness trainer for supporting exercise motivation? and 2) How can an AI-enabled fitness trainer be designed to enhance exercise motivation and promote sustained physical activity? To explore the topic in more detail, users’ ethical concerns towards AI, as well as goals for exercise, were also researched. The qualitative study was conducted through semi-structured user interviews with six volunteering participants.
The results show that participants preferred functionalities such as the ability to track and predict their fitness development with statistics and gather achievements from completed workouts, aligning with the persuasive design principles of “self-monitoring” and “rewards.” Seeing the sources of the AI-enabled fitness trainer’s recommendations was also highly valued, relating to the principle of “credibility.”
Less preferred or unmotivating persuasive functionalities were “reminders” about daily exercise, as well as “social comparison,” especially when not perceived as lighthearted fun within a friend group. Participants were divided between those who preferred exercising alone and those who gained motivation from social factors such as peer pressure and shared activity challenges.
Most proposed AI-enabled fitness trainer functionalities had both positive and negative feedback from the participants, indicating that motivation is highly based on personal preference. Environmental context, current energy levels, and busyness could also hinder the motivation to exercise. From this, it can be inferred that the AI-enabled fitness trainer should adapt to varying situations and preferences, thus offering highly personalized support.
Based on the findings, collectively preferred functionalities were combined into a table as design guidelines for the AI-enabled fitness trainer. Additionally, UI wireframes to visualize examples of design decisions based on these guidelines were implemented using Figma. The UI designs highlight persuasive elements such as modifiable exercise programs, rewards, and positive feedback.
This thesis offers a new perspective on combining persuasive design principles with AI-enabled fitness trainers. By better understanding the connection between persuasiveness and motivation in AI-enabled trainers, users can be better supported in progressing toward their individual fitness goals—whether that involves walking up stairs without losing breath or running a marathon.