Bio-generative art & music

01Overview and value chain#

Markers EC: no dedicated EU framework; falls under general consumer-product and data-protection rules | OECD: Cross-cutting bio×non-bio intersections & genomics and bioinformatics | Regulator: FTC (US, consumer product claims), EDPB (EU, data handling for biodata-driven installations)

Bio-generative art and music turns measurements from living systems — plant bioelectric signals, tree sap flow, ecological datasets, biological imaging — into music or immersive visual art rather than treating biology as raw material or subject matter alone. Data Garden’s PlantWave device measures changing electrical conductivity across a plant’s leaf surface, maps that “wave” into musical parameters such as pitch, and drives an instrument collection called a soundset, positioned as the direct successor to the earlier MIDI Sprout device with over a decade of engineering behind the current mapping pipeline. Refik Anadol Studio’s “Biophilia” series — including “Living Architecture: Biophilia” at the 2025 Venice Biennale and “Machine Dreams – Biophilia” in Mountain View, California — applies large-scale machine learning to biological and natural datasets to generate immersive architectural installations. Marshmallow Laser Feast’s “Of the Oak” (commissioned by the Royal Botanic Gardens Kew) and “Treehugger: Wawona” translate real sensor data on a tree’s internal water circulation and seasonal biological activity into multisensory audiovisual installations, treating the organism’s own physiological data as the composition’s source material. Genspace, a New York community biology lab founded in 2010, runs a five-month Artist-in-Residence program supporting artists whose work involves hands-on biological investigation, functioning as infrastructure for the wider bio-art field rather than a studio producing finished works itself.

The key directions of bio-generative art and music are:

1. Plant/fungal bioelectric sonification hardware: consumer devices measuring electrical activity in living plants and fungi, mapping the signal into real-time musical output.
2. AI-driven installations on biological datasets: large-scale generative art built from biological and natural-system datasets processed through machine-learning models, exhibited at museum and biennale scale.
3. Sensor-driven organism data as composition source: multisensory installations that sonify or visualize real physiological data from a specific living organism (tree sap flow, root networks) rather than abstract biological imagery.
4. Bio-art infrastructure and residencies: community labs and residency programs providing artists with lab access and biological-investigation training, supporting the field’s production pipeline upstream of finished artworks.

Sectoral value chain#

[Biological signal source: plant/fungal bioelectric activity, sap flow sensors, biological dataset] ──> [Signal capture / dataset acquisition] ──> [Mapping or model processing]
                                                                                                                                                                 │
                                                                                                                                                 (Real-time or rendered output generation)
                                                                                                                                                                 │
                                                                                                                                                                 ▼
[Finished output: music, installation, exhibition] <─── [Presentation format: device, gallery, museum commission] <─────┘

Value chain levels#

Cross-cutting technologies of the sector:

• Plant bioelectric signal sonification: electrode-based measurement of leaf-surface conductivity mapped into real-time musical pitch and rhythm parameters.
• Biological dataset generative AI: machine-learning models trained on or processing biological/natural-system datasets to generate large-scale visual and audio installations.
• Environmental biodata visualization: sensor-driven capture of an organism’s internal physiological activity (sap flow, root networks) rendered as multisensory audiovisual content.

02US#

The US hosts both ends of the sector — a commercial hardware company selling plant-music devices direct to consumers, a globally-commissioned AI-art studio working at architectural scale, and a community lab supporting the field’s artist pipeline.

plant bioelectric sonification devices, large-scale AI installations on biological data, bio-art residency infrastructure#

• Data Garden / PlantWave (US): sells the PlantWave device, successor to the earlier MIDI Sprout, which measures plant and fungal electrical activity and converts it into real-time music via mapped “soundsets”; supported EarthPercent at SXSW 2026.
• Refik Anadol Studio (Los Angeles, CA): produces large-scale machine-learning installations built on biological and natural datasets under its “Biophilia” series, commissioned for the 2025 Venice Biennale (19th International Architecture Exhibition) and exhibited in Mountain View, California in December 2025.
• Genspace (New York, NY): community biology lab founded in 2010 running a five-month Artist-in-Residence program for artists working with hands-on biological investigation, functioning as field-level infrastructure rather than a producing studio.

03CN#

No China-headquartered organization cleared this screening round for bio-generative art and music specifically. Chinese-language searches surfaced general AI-art festival coverage and academic art-history course listings rather than a vendor-own confirmation of a Chinese studio or company working specifically with biological data as a generative-art or music source.

screening limitation#

• Search results returned general AI-art event and course content rather than organization-specific evidence of a Chinese bio-generative art or music vendor as of this screening round.

04EU#

The EU’s contribution is anchored by the UK’s Marshmallow Laser Feast, whose installations translate a specific organism’s own sensor data — sap flow, seasonal rhythm — into audiovisual works commissioned by major cultural institutions.

sensor-driven organism data installations, cultural-institution commissions, multisensory mixed-reality format#

• Marshmallow Laser Feast (London, UK): artist collective founded in 2009; “Of the Oak” (2025, commissioned by the Royal Botanic Gardens Kew) visualizes an oak tree’s seasonal biological activity through interactive video, multichannel audio and an online field guide, while “Treehugger: Wawona” renders a giant sequoia’s internal water circulation as a multisensory mixed-reality installation.

05Leading companies and research institutes#

06Tech stack and innovations#

The sector’s stack spans three distinct approaches to turning biology into art — direct electrical sonification, dataset-driven machine learning, and organism-specific sensor networks — each aimed at a different presentation scale.

1. Bioelectric sonification (Data Garden/PlantWave):
   • Electrodes measure changing electrical conductivity across a plant or fungal surface, producing a continuous time-varying signal.
   • The signal is mapped directly into musical parameters (primarily pitch), driving pre-built instrument “soundsets” in real time without a machine-learning intermediary.
2. Biological-dataset generative AI (Refik Anadol Studio):
   • Natural and biological datasets are processed through machine-learning models to generate architectural-scale moving-image and spatial-audio content.
   • Installations are built for specific high-profile commissions (Venice Biennale, corporate campus exhibitions) rather than sold as a consumer product.
3. Organism-specific sensor networks (Marshmallow Laser Feast):
   • Real sensor data captured from a single living organism (a named oak, a specific sequoia) — sap flow, seasonal rhythm — becomes the direct source material for a multisensory installation.
   • Presentation format combines real-time video, spatial audio and a companion web field guide, commissioned by botanical and cultural institutions.

07Value chains and production pipelines#

Industrial pipeline of plant-bioelectric-signal music devices (IEC 60950 general consumer-electronics safety reference)#

┌───────────────────────────┐      ┌───────────────────────────┐
│ 1. Electrode attachment   │ ───> │ 2. Bioelectric signal capture│
└───────────────────────────┘      └───────────────────────────┘
                                                 │
                                                 ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 4. Soundset selection      │ <─── │ 3. Signal-to-pitch mapping│
└───────────────────────────┘      └───────────────────────────┘
              │
              ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 5. Real-time music synthesis│ ───> │ 6. Listener playback       │
└───────────────────────────┘      └───────────────────────────┘

Stage 1: Electrode attachment

A conductivity electrode is clipped to a plant leaf or fungal surface, establishing the physical connection to the living signal source.

Stage 2: Bioelectric signal capture

The device continuously reads the plant’s changing electrical conductivity, producing the raw time-varying “wave” that Data Garden’s mapping pipeline is built around.

Stage 3: Signal-to-pitch mapping

The captured wave is translated into musical parameters, primarily pitch, following the mapping logic refined across the MIDI Sprout-to-PlantWave product evolution.

Stage 4: Soundset selection

The mapped signal drives a pre-built instrument collection (“soundset”) chosen by the user, determining the timbral character of the resulting music.

Stage 5: Real-time music synthesis

The device synthesizes continuous music output live from the ongoing plant signal, with no pre-recorded or looped content.

Stage 6: Listener playback

The generated music streams to the listener in real time, tied directly to the living plant’s ongoing physiological state rather than a fixed composition.