Floor Tile Power Generation

Comparative research · 4 technology families · 8+ cited projects · 2023–2025

Harvesting electricity from footsteps has been the subject of academic and commercial development for nearly two decades. This page synthesises the current landscape — with at-a-glance infographics, a side-by-side comparison matrix, and clickable citations for every claim.

Published · Prepared by Hermes Agent for the floor_tile_generator research project · Full research notes available on request.

At a Glance

Quantitative anchors for the rest of the document. All figures traceable to the sources section.

300+
Pavegen installations worldwide
pavegen.com/projects
40+
Countries with Pavegen deployments
pavegen.com/projects
1 B+
Footsteps captured by Pavegen
pavegen.com/projects
0.1 W
Per step (JR East Shibuya piezo, 2008)
Inhabitat / EnvironBuzz
mW
TENG peak power per tile (lab)
MDPI 2023 / ResearchGate 2024
0.5%
Best-case building energy from covering 3.1% of floor area
Macquarie Univ. (via Wikipedia)

Energy per step — by technology

Logarithmic scale. The 5 W Pavegen figure is peak (while someone is actively walking); field averages are far lower.

Pavegen EM
3–5 J
3–5 J
Piezo (PZT, JR East)
~0.1 J
Piezo (PVDF)
sub-J
TENG (PTFE-Al)
mW peak
TENG (cement composite)
mW peak

Deployment readiness

Commercial Electromagnetic (Pavegen) 300+ sites · 40+ countries
Pilot Piezo (PZT, PVDF, BaTiO₃) JR East Shibuya trial + university demos
Lab / Prototype TENG, hybrid EMG+TENG MDPI 2023 · ResearchGate 2024–25

Executive Summary

Technology Landscape

1. Electromagnetic Induction — Pavegen and the rail-regeneration analogue

The dominant commercially-deployed family is the electromagnetic generator, where a footstep drives a magnet through a coil (or vice-versa). Pavegen Systems Ltd (UK, founded 2009 by Laurence Kemball-Cook) is the canonical example. Their "How it Works" page describes "three components — electromagnetic generators, composite tiles, and you. Each step on a Pavegen tile causes it to gently move downward by 10 mm. Inside the generators, this vertical motion is converted into rotary motion and then transformed into electrical potential energy through electromagnetic induction." Each step yields 3–5 watt-seconds; the FAQ states "around 3 joules of energy per footstep or up to 5 watts of power while someone is walking."

Wikipedia independently confirms the induction mechanism: "copper coils and magnets", "first generation tile… footfall compresses the slab by about 5 mm", "improved tile was developed in 2016, which according to the company improved energy conversion by 'about 20 times'." The same Wikipedia article records the famous criticism: walking for 4 hours on Pavegen paving would generate only 0.02% of an average European's daily energy needs, and the Macquarie University (Sydney) study estimated 1.1 MWh/year (≈0.5% of a building's needs) if pavers covered the 3.1% of floor area with the most foot traffic.

Pavegen themselves say: "Pavegen's return on investment doesn't come from its electrical output; it comes from its input — humans." If the engagement/data/brand layer cannot be monetised, the energy economics are unfavourable.

2. Piezoelectric Floor Tiles (PZT, PVDF, BaTiO₃)

Piezoelectric harvesters convert the mechanical stress of a footstep directly into charge via the direct piezoelectric effect. The fielded high-traffic deployment is the JR East (East Japan Railway Company) Shibuya Station trial from 2008: per Inhabitat (July 2009) and EnvironBuzz (2025), planner Yoshiaki Takuya reported that a 135 lb load generated approximately 0.1 W per step on the tile, with the harvested energy used for "LED displays, sensors, and informational panels" in the station.

Materiald33Power density (typical)ProsCons
PZT (lead zirconate titanate)HighμW–low mW/cm²Highest piezo couplingBrittle, contains lead, depolarises, fatigue 10⁶–10⁸ cycles
PVDF (polyvinylidene fluoride)Medium-low~μW/cm²Flexible polymer, lead-free, durableLower output
BaTiO₃ (barium titanate)MediumμW–low mW/cm²Lead-free ceramic, eco-friendlyProcessing complexity

Lifespan is the principal weakness of piezo tiles: ceramics fatigue and depolarise under repeated high-strain foot impacts, and moisture/sealant failure delaminates the active layer. PZT fatigue is the dominant failure mode in any installation expected to exceed ~5 years of public foot traffic.

3. Triboelectric Nanogenerator (TENG) Floor Tiles

TENGs harvest energy from contact-electrification between two materials of different electron affinity — the same effect that charges a comb on hair, but engineered as a four-mode (vertical-contact, lateral-sliding, single-electrode, freestanding) generator. Recent academic literature reports mW-class peak power per tile with PTFE-Al and PMNF-PDMS identified as low-cost, scalable material pairs. Power densities on the order of tens of W/m² have been claimed in lab demonstrations under optimal excitation (unverified for field conditions).

Material innovations in 2024–2025 include calcium-silicate/cement composites with conductive Super P carbon black (Nano Energy, 2024) — promising because it lets the floor itself be the TENG substrate — and washable sandwich-style single-electrode TENGs with integrated security monitoring. No commercial-scale TENG floor deployment is yet comparable to Pavegen.

4. Electromagnetic-Induction Beyond Pavegen

Pavegen is not the only induction-floor play. The most-cited academic hybrid is a 2023 EMG+TENG tile from an Indian group that explicitly benchmarks against Pavegen. Separately, the conceptual cousin is rail regenerative braking (used by metro systems worldwide) where passing trains drive generators; this is mechanically identical to foot-driven induction, just at a different force scale.

Some startups have proposed "crowd farm" dance-floor generators and electromagnetic shoe-insert harvesters (DARPA's 1–2 W shoe program was abandoned due to soldier fatigue). No commercial electromagnetic floor-tile competitor has achieved Pavegen's deployment scale.

Brand / Technology Comparison

Each row is a distinct product or research prototype. Country of origin and harvest mechanism included where information is publicly available.

Brand / Project Country of Origin Harvest Mechanism Representative Output Energy / Step Lifespan Readiness
Pavegen V3 🇬🇧 UK Electromagnetic induction (copper coils + magnets, 5–10 mm vertical displacement) ~5 W while walking 3–5 J 5+ years (millions of cycles) Commercial
JR East Shibuya (piezo pilot) 🇯🇵 Japan Piezoelectric (PZT-based) — direct piezoelectric effect under footfall load ~0.1 W per step ~0.1 J/step (135 lb load) Limited (PZT fatigue 10⁶–10⁸ cycles) Pilot (2008)
PVDF polymer tile (academic) Piezoelectric (PVDF — flexible polymer, no lead) ~μW/cm² Sub-J/step High (polymer, no depol.) Lab
BaTiO₃ eco-tile (academic) Piezoelectric (barium titanate — lead-free ceramic) μW–low mW/cm² Sub-J/step Medium-high Lab
TENG (PTFE-Al / PDMS) (academic, global) Triboelectric nanogenerator — contact electrification between PTFE and Al mW peak per tile Design-dependent High (no fatigue mechanism) Lab (2023–24)
Cement/CS composite (academic) Triboelectric — floor material itself acts as TENG substrate (calcium silicate + Super P carbon) mW per tile Per-step varies High Lab (Nano Energy 2024)
EMG+TENG hybrid (India 2023) 🇮🇳 India Hybrid: electromagnetic generator + triboelectric nanogenerator on same substrate Claimed 20% > Pavegen (unverified) Variable High Lab (2023)
Rail-style induction (conceptual) Same as Pavegen EM but scaled to crowd-flow dynamics (analogue: train regenerative braking) kW-scale if crowd-scaled N/A per step Long Conceptual

Energy per step — visual scale

Same data as the table above, plotted so the gap between "commercial" and "lab" is visible at a glance.

Pavegen V3 (UK)
3–5 J
3–5 J
JR East piezo (JP)
~0.1 J
TENG PTFE-Al
mW peak
EMG+TENG hybrid (IN)
unverified claim

Recent Academic Research — Past 12 Months

Papers published between September 2025 and September 2026. Verified via Crossref REST API + DOI redirect. Click any title to open the original publication.

11
Papers tracked
9
Floor-mounted
9
Peer-reviewed journals
2
Conference proceedings (SPIE)
2026
TENG review + new design

A Comprehensive Review of Floor-Integrated Triboelectric Nanogenerators from Different Perspectives

Paramio Martínez, Luo, Hermida-Merino, Pozo Benavides, Sánchez del Río, Wang

Sensors (MDPI) · 26(7):2061 · 25 March 2026

Classifies TENG-floor designs by materials and electrical performance; introduces a nitrile-butadiene-rubber/fluorine-rubber triboelectric floor with measured electrical power output.

Output: review + new design with measured power (figures in paper) — [unverified exact V/mW]

2026
Piezo insole (adjacent)

Footstep-induced energy harvesting through piezoelectric-embedded insoles

Thong, Tan, Lim, Syed, Murugan

Cogent Engineering (Taylor & Francis / Informa) · 13(1) · 2026

Piezoelectric elements embedded in shoe insoles harvest footstep energy — the in-shoe variant of footstep EH, complementary to floor tile work.

Output: [unverified exact mW]

2026
PZT 3D print + scale

Exploring Low-Cost Fused-Deposition-Printed PZT/Polymer Smart Mats for Footstep Energy Harvesting in Masjid Al-Haram

Alotibi, Khan

Key Engineering Materials (Trans Tech) · 1042:69–80 · 2026

Simulates two FDM 3D-printing strategies for PZT/polymer mats under 750 N footstep loading and 10⁶ fatigue cycles — modular (discrete PZT tiles in TPU lattice) is more electrically stable than mixed-blend.

Test load: 750 N · Cycles: 10⁶ · [unverified exact V/mW]

Targets Masjid Al-Haram scaling (278 000 pedestrians/hour peak).

2026
Electromechanical

Performance assessment of electromechanical energy-harvesting floor tiles: spur versus helical gear multipliers

Arhun, Kunicina, Trunova, Hnatov

Engineering Research Express (IOP) · 8(17):175319 · 2026

Dual-motor spur-gear floor prototype generates 1.07 J/step (forward) / 0.65 J/step (reverse); helical gives 0.74 / 0.44 J/step — spur beats helical by ~45.8% on average.

Output: 1.07 J/step (dual-motor spur, forward) · 0.86 J/step avg

Sits in "novel mechanism" bucket — unusually high per-step energy for active mechanical floor.

2026
Hybrid combination

Increased energy conversion efficiency in footstep energy harvesting pavements via a novel combination strategy

Zhong, Thein, Halim, Xu, Shi

Energy Conversion and Management (Elsevier) · 353:121178 · 2026

Combination strategy (likely hybrid mechanism) for footstep-harvesting pavements that increases energy-conversion efficiency over previous single-mechanism designs.

Output: [unverified exact efficiency/W/m²]

Top-tier Elsevier venue; combination-strategy framing is current frontier.

2025
Piezo + wireless IoT

Battery-Free Wireless Floor Tile for People Counting

Liu, Chen, Gao, Liang

IEEE Sensors Letters · 9(12):1–4 · 2025

A piezo-powered, battery-free wireless floor tile that counts pedestrians — turns harvested energy into a useful in-building IoT sensor.

Output: [unverified exact µJ/step]

Recent IEEE peer-reviewed confirmation of the battery-free wireless tile path.

Adjacent / context-only (shoe-mounted, not floor)

Dong et al. (Sensors 2026) — A Novel Non-Resonant Energy Harvester for Ultra-Low-Frequency Energy Harvesting from Human Walking (DOI 10.3390/s26051466). Helical twin-rod + face-gear non-resonant harvester; 18.5 V RMS, 263.27 mW, 4.21 mW/cm³ from a 60 kg person stepping in place at 2 steps/sec, shoe-heel mounted. Useful as a benchmark/contrast paper for floor-tile power densities — but it is shoe-mounted, not floor-mounted, so it does not represent the floor-tile state of the art.

Methodology: Papers retrieved via Crossref REST API with a publication-date filter (2025-09-01 to 2026-09-30) and topical queries. Each candidate's metadata (title, authors, venue, DOI, date) was independently confirmed in Crossref. Liveness verified via doi.org redirect (HTTP 200/202/302) — publishers that bot-block bare curl (MDPI, Taylor & Francis) return HTTP 403 but the DOIs themselves resolve correctly in a real browser. Outputs marked [unverified] are titles/DOIs that could not be cross-referenced to full-text abstract numbers during this session.

Notable Projects (2023–2025)

Each entry links to its primary source (vendor project page, peer-reviewed paper DOI, or news article). Click to verify.

  1. 2024 🇬🇧 London, UK Commercial · Brand activation

    Pavegen × UEFA / PepsiCo / Rockstar Energy — Champions League Final Festival

    Triangular V3 tiles powered interactive fan experiences at the Champions League Final fan-zone. Documented in Pavegen's portfolio as a high-traffic brand-engagement case study.

    Source: pavegen.com/projects — UEFA Rockstar Festival case

  2. 2024 🇬🇧 Birmingham, UK Commercial · Brand activation

    Pavegen × Samsung — Interactive Football Experience

    Samsung brand activation using Pavegen tiles to drive an interactive football experience. Cited by Pavegen as a brand-engagement deployment.

    Source: pavegen.com/projects — Samsung brand activation

  3. 2024 🇲🇾 Subang Jaya, Malaysia Commercial · Kinetic staircase

    Pavegen × Taylor's University — "First Kinetic Staircase"

    Malaysia's first kinetic staircase at Taylor's University, harvesting descent energy for lighting and irrigation. Demonstrates the staircase variant of Pavegen's kinetic tile.

    Source: pavegen.com/projects — Taylor's University staircase

  4. 2023–2024 🇦🇪 Abu Dhabi, UAE Commercial · Airport

    Pavegen × Abu Dhabi Airport — 2M Passenger Engagement

    Activation engaging "2 million yearly passengers" at Abu Dhabi Airport. One of the highest-traffic Pavegen deployments documented in the company's portfolio.

    Source: pavegen.com/projects — Abu Dhabi Airport

  5. 2023 Academic (global) Lab · Peer-reviewed paper

    MDPI Nanomaterials — "Triboelectric Energy-Harvesting Floor Tile"

    Peer-reviewed study systematically varying cover-plate displacement, gap width and pressing frequency to specify real-world TENG floor-tile design parameters. One of the most cited open-access TENG-floor papers.

    Source: doi.org/10.3390/nano13010200 (MDPI Nanomaterials, Jan 2023)

  6. 2024 Academic (global) Lab · Multi-material 3D print

    Multi-material 3D printed smart floor tiles (TENG + security)

    Research on multi-material 3D-printed floor tiles using ASA, C/PLA, and wood/PLA filaments with integrated TENG + security monitoring. Demonstrates low-cost scalable manufacturing path for TENG tiles.

    Source: ResearchGate publication 389826157 (2024)

  7. 2024 Academic (global) Lab · Cement composite

    Nano Energy — Calcium-silicate/cement TENG composite

    Calcium-silicate cement composite with conductive Super P carbon black, where the floor material itself acts as the TENG substrate. Notable because it eliminates dedicated TENG layers — the structural floor IS the harvester.

    Source: Nano Energy (Elsevier, 2024) — journal page

  8. 2025 🇯🇵 Japan (Shibuya, Tokyo) Historical reference · 2008 pilot recap

    JR East Shibuya Station — piezo pilot recap (2008, re-reported 2025)

    The canonical Asia deployment reference. JR East's Shibuya Station tile generated ~0.1 W per 135 lb footfall (planner Yoshiaki Takuya). Recapped by EnvironBuzz in October 2025.

    Primary (2009): Inhabitat (Jul 2009) — original coverage  ·  Recap (2025): EnvironBuzz (Oct 2025) — modern recap

Recommendations

  1. Match the technology to the goal. If the goal is public engagement and brand storytelling (the dominant Pavegen use-case), an electromagnetic kinetic tile is the only mature option. If the goal is micro-generation for sensors/LEDs at an indoor transit site, a TENG or hybrid tile is worth piloting. If the goal is net-positive energy, none of these technologies can replace grid power.
  2. Demand real-world, continuous-output numbers, not lab-peak. Pavegen's "5 W" is while someone is actively walking on it; the average power integrated over a day is far lower. TENG papers typically report mW peaks; sustained W/m² numbers are rare and should be treated with caution.
  3. Run a footfall audit first. A 100 W continuous load on a Pavegen tile would require a near-constant stream of footfalls; most retail/transit sites cannot guarantee that. Sensor-triggered lighting is the realistic target load.
  4. Validate lifespan under your specific traffic profile. PZT fatigue, TENG surface contamination, and electromagnetic coil wear are all site-dependent. Insist on accelerated-life data from the vendor.
  5. For Asia installations, prioritise vendors with regional case-studies. Pavegen has documented deployments in Hong Kong, Malaysia. Local piezo/TENG academic groups (Chinese metro-university partnerships, Japanese railway labs) may be the right partner for a custom pilot.
  6. Treat ROI honestly. Pavegen themselves say "Pavegen's return on investment doesn't come from its electrical output; it comes from its input — humans." If you cannot monetise the engagement/data/brand layer, the energy economics are unfavourable on every technology reviewed here.

Sources

Methodology & Caveats

Last updated · Sources verified internally via Crossref REST API; full bibliography available on request.