Adaptive Self-Efficacy-Based AI Coaching for Cycling (AI)

Adaptive Self-Efficacy-Based AI Coaching for Enhanced Indoor Cycling Performance: A Personalized Machine Learning Approach

The primary objective of this study is to evaluate whether adaptive, AI-delivered personalized self-efficacy-based AI coaching based on real-time physiological and performance feedback enhance indoor cycling power output during a 20-minute time trial compared to static affirmations and exercise-only control conditions.

Study Overview

• Status: Not yet recruiting
• Conditions: Exercise Training; Motivation for Physical Activity; Motivational Enhancement; Exercise Behavior; Exercise Adherence Challenges
• Intervention / Treatment: Behavioral: Group 1: Self-efficacy-based AI coaching; Behavioral: Group 2: Static AI Affirmations

• Study Type: Interventional
• Enrollment (Estimated): 120
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Anna Queiroz, Ph.D.; Phone Number: 305-284-3752; Email: aqueiroz@miami.edu
• Study Contact Backup: Name: Meshak Cole, B.S.; Phone Number: 305-284-3752; Email: mwc94@miami.edu

Study Locations

• United States
  • Florida
    • Coral Gables, Florida, United States, 33146
      • University of Miami
      • Contact: Anna Queiroz, Ph.D.; Phone Number: 305-284-3752; Email: aqueiroz@miami.edu
      • Contact: Meshak Cole, B.S.; Phone Number: 305 2843752; Email: mwc94@miami.edu

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• Age 18-40 years

  • Recreationally active
  • Familiar with stationary cycling
  • Able to complete 20 minutes of vigorous cycling

Exclusion Criteria:

• Cardiovascular, metabolic, or respiratory conditions

  • Medications affecting heart rate response
  • Lower extremity injury within past 3 months
  • Competitive cyclists (>10 hours cycling/week)
  • Pregnancy

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Basic Science
• Allocation: Randomized
• Interventional Model: Parallel Assignment
• Masking: Single

Number of Arms

3

Arms and Interventions

Participant Group / Arm	Intervention / Treatment
No Intervention: Control Group; No affirmations delivered. Participants receive only time notifications at 5, 10, 15, and 19 minutes for pacing awareness. Same equipment worn to control for potential monitoring effects.
Experimental: Group 1: Self-efficacy-based AI coaching; The Thompson Sampling contextual bandit algorithm, trained on Session 1 data, monitors performance continuously and evaluates every 5 seconds whether to deliver an affirmation.	Behavioral: Group 1: Self-efficacy-based AI coaching; The Thompson Sampling contextual bandit algorithm, trained on Session 1 data, monitors performance continuously and evaluates every 5 seconds whether to deliver an affirmation. The policy is trained to maximize a multi-objective "efficacy-preserving performance" function that rewards: Maintaining target power relative to rolling 30s/2min/5min baselines; Stabilizing short-horizon power variability (30s coefficient of variation); Stabilizing heart-rate (HR) trajectory consistent with efficient pacing The decision process considers: Current power relative to 30-second, 2-minute, and 5-minute rolling averages; Power output variability (coefficient of variation over past 30 seconds); Heart rate trajectory and cardiac drift patterns; Cadence stability and changes from baseline; Time elapsed and expected fatigue progression based on power-duration curve Self-efficacy-based AI coaching adapts to physiological measures (power and heart rate).
Active Comparator: Group 2: Static AI Affirmations; Generic motivational messages delivered at fixed intervals (minutes 3, 6, 9, 12, 15, and 18) regardless of performance state. Messages follow the same complexity gradient based on elapsed time rather than individual response.	Behavioral: Group 2: Static AI Affirmations; Generic motivational messages delivered at fixed intervals (minutes 3, 6, 9, 12, 15, and 18) regardless of performance state. Messages follow the same complexity gradient based on elapsed time rather than individual response: Minutes 3, 6: "You're building momentum with every pedal stroke-maintain this strong rhythm"; Minutes 9, 12: "Strong effort-push through this challenge"; Minutes 15, 18: "Final push-finish strong"

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Mean cycling power output during 20-minute time trial	Average cycling power output over the full 20-minute time trial. The outcome compares mean power between intervention arms (adaptive AI coaching vs. static affirmations vs. exercise-only control). Power is captured continuously via the cycling ergometer and summarized as the mean watts for each participant's trial.	Day 2

Collaborators and Investigators

• Sponsor: University of Miami
• Investigators: Principal Investigator: Anna Queiroz, Ph.D., University of Miami

Study record dates

Study Major Dates

• Study Start (Estimated): June 1, 2026
• Primary Completion (Estimated): December 23, 2028
• Study Completion (Estimated): December 28, 2028

Study Registration Dates

• First Submitted: December 19, 2025
• First Submitted That Met QC Criteria: December 22, 2025
• First Posted (Actual): January 5, 2026

Study Record Updates

• Last Update Posted (Actual): March 3, 2026
• Last Update Submitted That Met QC Criteria: February 27, 2026
• Last Verified: February 1, 2026

More Information

Terms related to this study

• Keywords: Physical Activity; cycling; just-in-time adaptive intervention; AI affirmations
• Additional Relevant MeSH Terms: Behavior; Motor Activity
• Other Study ID Numbers: 20251354

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No