Tuning pianos using reinforcement learning

Matthew Millard, Hamid R. Tizhoosh

Research output: Contribution to journalArticlepeer-review


The tuning system of a piano has remained relatively unchanged since the instrument's inception. A piano's tuning system has been designed to be both inexpensive to manufacture and to preserve the tension and thus pitch of each string over long periods of time. This tuning system requires such a high degree of skill to manipulate that only trained professionals are able to tune pianos. This paper presents a novel adjustable impact tuning hammer and a reinforcement learning control system that may allow piano owners to tune their own pianos in the future.

Original languageEnglish (US)
Pages (from-to)576-593
Number of pages18
JournalApplied Acoustics
Issue number5
StatePublished - May 2007


  • Automated piano tuning system
  • Impact tuning hammer
  • Piano tuning
  • Reinforcement learning

ASJC Scopus subject areas

  • Acoustics and Ultrasonics


Dive into the research topics of 'Tuning pianos using reinforcement learning'. Together they form a unique fingerprint.

Cite this