PaperarXiv:2610.40370
The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)
arXiv 2610.40370v1 (submitted 2026-10-29, cs.RO primary; arXiv licence CC BY-NC-ND 4.0)
(opens in a new tab)Moradi, Behnam, "The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables", arXiv:2610.40370, CC BY-NC-ND 4.0
TL;DR
It treats coaching as accountable sequential experimentation. The agent proposes the next experiment by information-directed Bayesian design and checks feasibility with a safety officer. Underneath is a wrist-only mechanistic digital twin for mountain biking, built from three parts: a lagged heart-rate response model on repeated climbs; a latent descending-skill state; a fitness–fatigue model with learned time constants. The agent is built on Google ADK with Gemini; the repo default is gemini-3.8-flash.
- Form factor: wrist
Abstract
Large language model (LLM) agents for personal health read wearable data and recommend, but do not learn about the individual in a way that can be audited or falsified. We present EpiCoach, an agentic system that treats one athlete, a cross-country mountain biker, as the subject of accountable N-of-1 science. Claims about the athlete live in an evidence-gated hypothesis registry: each must issue a probabilistic prediction about a future session before it occurs, and its belief moves only through computed evidence from a logged statistical fit. Predictions come from a mechanistic digital twin inferred from heart rate, GPS and barometric altitude: a lagged heart-rate response model, a fitness-fatigue model with learned time constants, and a latent technical-descending skill state, measured on discovered trail segments. The agent prescribes the next training week as a constrained experiment maximizing information gain subject to load and safety constraints, around two instrument rides. We evaluate the system two ways over eighteen months of data. A rolling-origin replay benchmark shows the twin beats standard forecasters on heart rate but not segment time, and the agentic loop matches a static model. A live prospective phase runs the weekly cycle: in its first three weeks the ledger sealed 20 predictions, scored 14, the descent forecast calibrated after one correction, the climb forecast pessimistic as the athlete improved, and the safety officer vetoed the agent twice. Every correction is preserved in the registry, a template for auditable, falsifiable personalization in digital health. We release the registry, benchmark and protocol.
Verbatim from arXiv (opens in a new tab) · arXiv metadata (abstracts are released under CC0)
Results on ChatPPG leaderboards
All leaderboardsNo results from this paper on ChatPPG leaderboards yet. Browse the leaderboards.
Tasks & signals
Tasks
Signals
Keywords
How it works
Cite this paper
Moradi, B. (2026). The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach). arXiv 2610.40370v1. https://arxiv.org/abs/2610.40370
Moradi, Behnam. 2026. "The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)." arXiv 2610.40370v1. https://arxiv.org/abs/2610.40370.
Moradi, B. (2026) 'The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)', arXiv 2610.40370v1. Available at: https://arxiv.org/abs/2610.40370
Moradi, Behnam "The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)." arXiv 2610.40370v1, 2026. https://arxiv.org/abs/2610.40370.
B. Moradi, "The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)," arXiv 2610.40370v1, 2026, [Online]. Available: https://arxiv.org/abs/2610.40370.
@misc{moradi2026athlete,
title = {{The Athlete as a Scientist of One: Agentic Bayesian Experimental Design for Individualized Physiology from Consumer Wearables (EpiCoach)}},
author = {Moradi, Behnam},
year = {2026},
howpublished = {arXiv 2610.40370v1},
eprint = {2610.40370},
archivePrefix = {arXiv}
}Added to the ChatPPG research hub
