Modulor

Confidential Material

This document contains proprietary information about Modulor, including product design, technical specifications, intellectual property, and business strategy. Please identify yourself and accept the terms below before proceeding.

Non-Disclosure Agreement

By accessing this document, you ("Recipient") agree to the following terms:

1. Confidentiality. All information contained in this document — including but not limited to product designs, technical specifications, engineering details, patent claims, market analysis, pricing, and business strategy — is confidential and proprietary to Modulor ("Disclosing Party").

2. Non-Disclosure. Recipient agrees not to disclose, publish, or disseminate any confidential information to any third party without prior written consent from the Disclosing Party.

3. Non-Use. Recipient agrees not to use any confidential information for any purpose other than evaluating a potential business relationship with Modulor.

4. No Copies. Recipient shall not copy, reproduce, or distribute this document or any portion thereof without written permission.

5. Return of Materials. Upon request, Recipient shall return or destroy all copies of confidential materials.

6. Duration. This agreement remains in effect for a period of two (2) years from the date of acceptance.

7. No License. Nothing in this document grants any license or rights to any intellectual property owned by Modulor.

This NDA is governed by the laws of the State of Delaware. By checking the box below, you acknowledge that you have read, understand, and agree to be bound by these terms.
← Back to home (decline)
← Home

MODULOR

Actuarial infrastructure for
musculoskeletal injury risk.

The hardware is the data layer. The data is the business.

Confidential — For potential partners only

1

The Problem

Soft-tissue injuries are the #1 cause of lost games in elite sports — and they're getting worse.

Strength has force plates. Speed has timing gates. Recovery has Normatec. But flexibility and joint mobility — the biggest levers for injury prevention — are still estimated by eye and documented on paper.

The full scope — every muscle group Modulor addresses:

• Muscle strains are the largest single injury category in elite football: hamstrings 42%, adductors 25%, quadriceps 19%, calves 14%. Re-injury rates: adductors 29%, hamstrings 30%. (Eirale et al., epidemiological review, Br J Sports Med)

• 25% of NFL players suffer a lower-extremity strain every season — 33% re-injury rate. (Jenkins et al., 12-year review of 2,101 NFL hamstring injuries, Orthop J Sports Med 2024)

• 21% of professional footballers suffer a groin/adductor injury every season; 68% are adductor-related. (Werner et al., 17-club 606-player UEFA prospective study, Br J Sports Med)

• Restricted hip internal rotation and ankle dorsiflexion are independent risk factors for ACL injury — 83% of studies in a 2,819-athlete systematic review confirmed the link. (Belkhelladi, Cierson & Martineau, Orthop J Sports Med 2025)

• Hamstring injuries have DOUBLED from 12% to 24% of all injuries in men's professional football over 21 years — training injuries rising 6.7% annually despite modern training science. (Ekstrand et al., UEFA Elite Club Injury Study, Br J Sports Med 2023)

• Economic impact: top LaLiga clubs lose €7M+ annually to muscle injuries; Premier League teams spent £266M total on sidelined players' wages in 2023/24 — £83M on hamstrings alone. (Melcón-Ballesteros et al., PLOS One 2024; Premier Injuries Ltd. 2024)

2

The Solution

Keep athletes on the field. Extend their careers. Sharpen their edge.

Modulor is a single-station platform that perfects an athlete's ability to compete — applying precise motorized cable tension while measuring range of motion, force response, and movement quality in real time.

3

The Product

Controlled tension. Measured range. Data on every rep.

FIG. 2 — Apparatus Side-View Schematic

BLDC 20:1 LOAD θ E 10 14 16 18 20 22 24 12 26

10 BLDC motor   12 Planetary gearbox   14 Cable spool   16 Inline load cell   18 Quick-release cuff   20 Depth camera   22 Touchscreen   24 E-stop   26 Frame + bench

Modulor is a single-station platform that applies precise, motorized cable tension to stretch an athlete's muscles while simultaneously measuring range of motion, force response, and movement quality — all in real time.

One station delivers:

  • Standardized assisted stretching — every athlete, every session
  • Precise joint-angle measurement (±2° accuracy)
  • Objective, exportable flexibility documentation
  • Protocol consistency across the entire roster
  • Longitudinal data that identifies injury risk early
4

How It Works

Eight minutes. Nine variables. Zero guesswork.

FIG. 3 — Stretch Session Method (Bilateral Assessment Protocol)

STEP 100 BADGE IN Load athlete profile + session history + training context STEP 102a BASELINE — L Passive stretch @ 8 lb Measure peak ROM (θ_L) STEP 102b BASELINE — R Passive stretch @ 8 lb Measure peak ROM (θ_R) STEP 104 ASYMMETRY CALC Δ = |θ_L − θ_R| Flag if Δ > 10° STEP 106 STRETCH PROTOCOL Motor ramps cable PID @ 100Hz (FIG. 10) Vision tracks joints Compensation detect (FIG. 11) 100Hz closed loop STEP 108 DATA LOG ROM, force, asymmetry compensation flags tolerance threshold pain response binary STEP 110 REPORT Session Δ ROM Asymmetry trend Risk score update STEP 112 — Longitudinal data feeds next session's baseline targets and protocol parameters

Method claim: Steps 100–112 define the complete bilateral assessment and adaptive stretching protocol with longitudinal feedback loop.

5

Technology

Five systems. One closed loop.

FIG. 4 — Cable Drive Assembly

BLDC 200W 30 20:1 PLANETARY 32 ⊕ 34 CLUTCH 36 DYNEEMA CABLE 38 ODrive FOC 100Hz PWM

Cable Drive

200W BLDC motor, 20:1 planetary gearbox, Dyneema cable. ODrive FOC controller at 100Hz. Force-compliant, backdriveable.

FIG. 5 — Sensor Array & Data Paths

40 RealSense D435 LOAD CELL 42 Inline, 100Hz JETSON EDGE COMPUTE 44 30fps depth 100Hz force 15" TOUCH 46 CLOUD 48

Depth Camera + Sensors

Intel RealSense D435. 12-point skeletal tracking at 30fps. Inline load cell at 100Hz. Edge compute on Jetson module.

FIG. 6 — Control Interface Architecture

STAFF / OPERATOR LAYER SESSION UI Protocol select 50 LIVE DISPLAY Force + ROM 52 TREND VIEW Longitudinal Δ 54 LOCAL DATA BUS — Sensor fusion @ 100Hz → UI update @ 30Hz → Cloud sync @ session end

Live Interface

15" touchscreen. Real-time force graph, ROM tracking, bilateral comparison. Longitudinal trend view with export.

FIG. 7 — Five-Layer Safety Cascade

L1 SOFTWARE FORCE CAP — max tension limit, cannot be overridden by protocol L2 PID COMPLIANCE — force setpoint tracking, auto-reduce on overshoot L3 SLIP CLUTCH — mechanical torque limit, cable slips at threshold L4 E-STOP — hardwired motor kill, staff or athlete accessible L5 QUICK-RELEASE CUFF — instant mechanical disconnect <10ms <10ms passive <50ms instant Each layer operates independently — any single layer can halt the stretch

Five-Layer Safety

Software cap → PID compliance → slip clutch → E-stop → quick-release. Each layer is independent and fail-safe.

6

What It Measures

Nine variables. Measured, not estimated.

FIG. 8 — Measurement Extraction Pipeline (Raw Sensor → Derived Variables)

DEPTH CAMERA 640×480 @ 30fps LOAD CELL 100Hz analog MOTOR ENC Position + velocity POSE ESTIMATION 12-joint skeleton 3D coordinates FORCE PROCESSING Filtered, calibrated CABLE LENGTH SENSOR FUSION 100Hz sync 60 V1 Joint Angle (θ peak) V2 Cable Force (F peak, σ) V3 Bilateral Asymmetry (Δ) V4 Hold Time (t @ end-range) V5 Compensation Flags V6 Tolerance Threshold V7 Pain Response (binary) V8 Tension Ramp Rate V9 Session Δ (acute + chronic) RISK SCORE Longitudinal trend analysis → injury prediction 62

Apparatus claim: Three independent sensor inputs (40, 42, encoder) are fused at element 60 to derive nine biomechanical variables (V1–V9), fed to longitudinal risk scoring engine (62).

Joint Angle

Peak passive ROM at standardized force. ±2° accuracy. Comparable across sessions.

Cable Force

Inline load cell at 100Hz. Logs tension ramp, peak force, and variability during hold.

Asymmetry

Left vs. right at matched force. Flags differences >10%. Tracks resolution over time.

Hold Time

Duration at end-range without retreating. A neuromuscular tolerance metric.

Compensation

Camera detects pelvic rotation, knee bend, trunk shift. Quantified, not eyeballed.

Tolerance Threshold

The force level where the athlete first reports discomfort or resists. Tracked over weeks.

Pain Response

Binary flag per rep. Correlated with force and ROM to build a tissue-response profile.

Tension Ramp

Rate of force acceptance during initial stretch. Indicates tissue readiness and stiffness.

Session Δ

Acute ROM gain per session. Chronic trend over weeks. The metric that matters most.

7

Competitive Landscape

Modulor is the only platform that combines motorized stretching with real-time measurement.

Data-Rich / Measured No Data / Manual Passive / Manual Active / Motorized
Modulor
Biodex / Cybex
Normatec
StretchLab
Manual PT
Foam Rollers
Goniometer

Modulor is the only platform that occupies the upper-right quadrant — combining motorized, force-controlled stretching with real-time biomechanical measurement and longitudinal data tracking.

Biodex / Cybex

Clinical isokinetic. $100K+. Not designed for stretching protocols. No vision tracking.

StretchLab

Manual 1:1 service. $60/session. No measurement, no data, no consistency.

Normatec

Compression recovery. No active stretch. No ROM measurement.

Manual PT

Subjective assessment. No force control, no objective measurement. Rarely documented longitudinally.

8

Market Opportunity

Four buyers. One data asset.

Stretching is the wedge. The product is a measurement + risk-pricing layer. Four independent buyers monetize the same underlying corpus.

Buyer 1 · Hardware lease

Teams — pro + collegiate

~800 addressable programs in NA + EU. Target $60K/yr/station. Every franchise has a training-room budget and an injury-prevention mandate.

Buyer 2 · Data licensing

Insurers — WC + disability

$70B+ US market. MSK is ~40% of claims. Carriers buy aggregated risk signals, not hardware. Target: $250K–$1M per carrier / yr.

Buyer 3 · Enterprise

Leagues + PAs

Roster-level availability models. Standardized injury reporting aligned to CBAs. Long sales cycle — large check.

Buyer 4 · Strategic exit

Hyperice · Catapult · Nike

Patent + data corpus are the asset. Acquisition-optimized. Adjacent distribution already owns the consumer surface.

+ 5th pathway · DoD — non-dilutive

~800K MSK injuries/yr across active-duty — 100× pro sport. AFWERX + SOFWERX + DIU are parallel rails, not a pivot. See Doc 19 →

Hardware is the wedge

Lease-only model. Every install compounds the data corpus. Physical switching cost protects the moat once deployed.

9

Military / DoD Pathway

100× the injury volume. Non-dilutive capital. Parallel track.

MSK injury is the #1 cause of lost duty days across every service branch. ~800K/yr active-duty MSK injuries; $3.7B annual cost. We pursue this in parallel to commercial — funded by the government, not the seed.

Pathway 1 · 90-day yes

AFWERX · Air Force SBIR

Open Topic quarterly. $75K Phase I (3 mo) → $1.7M Phase II (9 mo) → $15M STRATFI. Target sponsor: PJ / combat controller / pilot training units.

Pathway 2 · Best-paying pilot

SOFWERX · USSOCOM

Tier-1 operators = closest analog to pro athletes. CRADA + funded pilot via Tech Experimentation event. $250K–$3M, 90–180 days.

Pathway 3 · Biggest check

DIU · Prototype OTA

Dual-use commercial-first. $1–20M prototype awards. Unlocked after first commercial pilot revenue. Target: Year 2.

Pathway 4 · Validation

DHA · CHAMP clinical

Peer-reviewed ROM/asymmetry → MSK injury prediction paper. Unlocks insurance licensing narrative. 6–24 mo, $500K–$10M R&D.

Product delta: minimal

Same hardware. Protocol library + unit-aggregated dashboards + IL-4 gov cloud path. ~2 person-weeks of integration per pathway.

What it does for the seed

"$3M + $500K non-dilutive Phase I" = smarter capital + diversified buyer + de-risked insurance narrative. Full plan in Doc 19.

9b

Product Roadmap

Honest scope. No overclaiming. PNF at V1.5.

FIG. 9 — Multi-Pattern Reconfigurable Platform (Single Frame, Seven Patterns)

SINGLE FRAME BLDC MOTOR (10) 34 P1 V1 FIXED P2 V2 SLIDING DEPTH CAMERA (40) V1 — LAUNCH (4 Patterns) P-A Hamstring SLR Supine, leg elevated. Cable at ankle. Straight-leg raise to end-range. P-B Adductor Abduction Supine, leg lateral. Cable pulls outward. Measures abduction ROM. P-C Hip Flexor Extension Prone/side-lying. Cable pulls into extension. Stretches hip flexors. P-D PNF Contract-Relax Applied to any pattern. Athlete pushes against cable (isometric), motor holds, then advances to new end-range. Motor detects force reversal. V2 — YEAR 2 (3 Patterns) P-E Hip Internal / External Rotation Seated, knee bent 90°. Cable at ankle. Rotational stretch. Requires P2 at 45°. P-F Ankle Dorsiflexion Seated, foot on plate. Cable at forefoot. Measures DF ROM. Requires P2 low. P-G Thoracic Rotation Seated, cable at shoulder harness. Rotational stretch. Requires P2 high. V1 → V2 UPGRADE PATH Same motor (10), same cable (38), same spool (34) Add: sliding track bracket + software update Hardware cost: ~$200 | Software: OTA update

Apparatus claim: A single motorized cable platform (elements 10, 34, 38) serves seven distinct stretch patterns through athlete repositioning (V1) and adjustable pulley position on a sliding track (V2), without changing the drive system.

V1 — Launch

Four stretch patterns

  • Hamstring SLR (straight-leg raise)
  • Adductor abduction
  • Hip flexor extension
  • PNF contract-relax overlay

Same motor, same cable, same pulley for all four. The athlete changes position — the machine stays simple.

V2 — Year 2

Three additional patterns

  • Hip internal / external rotation
  • Ankle dorsiflexion
  • Thoracic rotation

Requires adjustable pulley position on frame (sliding track). Same motor and cable system. Software update + hardware bracket.

10

Intellectual Property · Patent Pending

With a patent to protect it.

USPTO Provisional Application 64/040,028 · filed April 15, 2026 · 35 U.S.C. § 111(b)

FIG. 1 — System Block Diagram (Force-Controlled Stretch with Compensation-Aware Adjustment)

200 ATHLETE ID Badge + profile + session history 202 PROTOCOL ENGINE Pattern select Force target (F_set) 204 PID FORCE CONTROLLER e(t) = F_set − F_meas u(t) = Kp·e + Ki·∫e + Kd·ė → Motor PWM command 100Hz update rate 206 CABLE DRIVE BLDC + gearbox + spool + cable + slip clutch F_meas (load cell @ 100Hz) 208 ATHLETE On platform Limb stretched observes 210 VISION SYSTEM Skeleton + joint angles + compensation detect COMPENSATION SIGNAL → modifies F_set (NOVEL) 212 DATA ENGINE 9 variables / rep Longitudinal trends Asymmetry alerts Injury risk score (62) 214 — ADAPTIVE FEEDBACK: Historical data modifies next session's protocol parameters

System claim: Elements 200–214 define the complete closed-loop system. The red dashed path (210 → 204) represents the novel compensation-aware force adjustment — vision-detected compensation modifies the PID setpoint in real time.

FIG. 10 — PID Force-Compliance Control Loop (Element 204 Detail)

F_set (from protocol) Σ + − e(t) PID CONTROLLER Kp=2.4 Ki=0.8 Kd=0.15 Anti-windup enabled u(t) FORCE CAP clamp(u, F_max) MOTOR DRIVE ODrive FOC PLANT Cable + limb + tissue viscoelastic response F_meas (load cell) FORCE FEEDBACK PATH — 100Hz sample rate, <10ms loop latency COMP. SIGNAL (from 210) Reduces F_set when compensation detected

Method claim: The PID controller (204) maintains force setpoint while the compensation signal from the vision system (210) dynamically modifies the target, creating a dual-feedback architecture unique to this apparatus.

FIG. 11 — Compensation Detection & Adaptive Force Response (Element 210 Detail)

DEPTH FRAMES 640×480 @ 30fps 220 POSE MODEL 12-joint skeleton 3D coordinates (hip, knee, ankle, pelvis...) 222 ANGLE CALC θ = arctan2(Δy, Δx) per joint pair 224 COMPENSATION CLASSIFIER Pelvic rotation > 5°? Knee bend > 10°? Trunk lateral shift > 3cm? NO COMP → HOLD Maintain F_set COMP DETECTED F_set *= (1 − α·severity) α = 0.15 per threshold → PID (204) Modified F_set sent to motor LOG: comp type, severity, timestamp

Method claim: The compensation classifier (224) analyzes pose data to detect pelvic rotation, knee bend, or trunk shift, then proportionally reduces force setpoint — the core novel mechanism that distinguishes Modulor from all existing stretch devices.

Claim 1 — Method

A method for force-controlled stretching with compensation-aware adjustment, comprising: applying motorized cable tension via PID force control (204), detecting movement compensation via depth-camera pose estimation (210, 224), and dynamically modifying the force setpoint in response to detected compensation.

Claim 2 — Apparatus

An integrated cable-drive and vision feedback stretching apparatus comprising: a motorized spool (34) with inline force sensing (42), a depth camera (40) with skeletal tracking (220), a PID compliance controller (204), and a compensation classifier (224) that adjusts cable tension in real time.

Claim 3 — System

A bilateral assessment and adaptive protocol system comprising: standardized force application (F_set), bilateral ROM comparison (Δ = |θ_L − θ_R|), longitudinal trend analysis with injury risk scoring (62), and adaptive protocol modification (214) based on historical session data.

Figures 1–11 document the complete system. Prior art gap is strong — no existing device combines motorized cable stretching with real-time vision-based compensation detection and adaptive force modification.

11

Revenue Model

Multiple revenue streams, hardware-led foundation.

2026–2027 (Launch)

  • Hardware sales — 70%
  • SaaS subscriptions — 15%
  • Service / install — 15%

2028+ (Scale)

  • SaaS subscriptions — 45%
  • Hardware sales — 30%
  • Data + insights — 15%
  • Service / install — 10%
$15–25K
Per unit hardware
$500/mo
SaaS per machine
85%+
Gross margin on SaaS
12

Growth Engine

Modulor Flywheel: Compounding data advantage.

MODULOR
Athletes → Sessions, Data
Facilities → Adoption, Revenue
Data → Insights,
Protocols
Athletes: Sessions → Data → Better Protocols → Better Outcomes → Retention
Facilities: Adoption → Results → Reputation → Referrals → More Facilities
Data: Volume → Accuracy → Prediction → Injury Prevention → Value
Booster: Every prevented injury is a case study that sells the next 10 machines.
13

Integration Layer

We are the measurement layer. Every performance-stack pipe is a distribution lever.

We do not build our own AMS, programming app, or wearable. Every existing platform is a pipe we feed. That's how the hardware becomes infrastructure, not a sensor.

Tier 1 · V1 must-haves

  • Smartabase — #1 AMS in pro + elite college
  • Kitman Labs — EPL, NFL, MLS footprint
  • TeamBuildr — college S&C standard
  • Train Heroic — tactical + CrossFit delivery app
  • CSV + public REST API — universal fallback

Pilot gate: "we use X" cannot be a reason we lose the deal.

Tier 2 · 6-month unlock

  • Catapult — workload-to-MSK risk correlation
  • Polar Team Pro — collegiate workload
  • Hudl / Teamworks — roster + schedule
  • Epic / Cerner (HL7 FHIR) — medical + RTD

Each unlocks a specific buyer category.

The insurance play

Carriers don't integrate with point tools. They license aggregated, anonymized risk signals via data-licensing APIs. Target: $250K–$1M / carrier / yr after Year 2. No claims-system integration needed — just signed BAAs and a data feed.

The rule

Never write bespoke code for one customer. Every integration is a reusable template. Every custom request gets redirected to the public REST API + webhook pattern. Cheap per customer, compounding over time.

Ownership

Data-rights clause in every lease contract. Teams get their data back into their AMS; we retain the de-identified corpus in perpetuity. This is the insurance-licensing precondition. Full plan: Doc 20.

13b

Capital Requirements

Raising $3M Seed — strategic-led preferred.

30-month runway. 5–8 pilot stations deployed. Patent utility conversion. Path to Series A on $2–4M ARR + insurance LOI.

  • Hardware — prototype, pilot stations, tooling — 40% ($1.2M)
  • Engineering — firmware, CV/ML, full-stack — 30% ($900K)
  • GTM — pilot ops, clinical validation, partnerships — 15% ($450K)
  • IP — utility conversion, PCT, continuations — 10% ($300K)
  • Reserve — legal, ops, contingency — 5% ($150K)

ARR Projection by Installed Units

Hardware SaaS ($6K/yr/unit)
$260K
10
$650K
25
$1.3M
50
$2.6M
100

Active Units Installed

● Hardware

$1.2M
Mechanical design, CAD, CNC + frame fabrication$250K
Motor, gearbox, controller — 8 pilot stations$320K
Depth camera, load cell, compute, touchscreen × 8$180K
Tooling, fixtures, small-run production setup$200K
Iteration (v1.5 + PNF firmware rev)$150K
Shop time, assembly, QA$100K

● Engineering

$900K
Firmware engineer — PID loop, PNF detection, safety$220K
CV/ML engineer — compensation detection, pose fusion$260K
Full-stack — API, AMS connectors, dashboards$200K
Mechanical / EE fractional$150K
Cloud infra, data pipeline, security$70K

● GTM

$450K
Pilot operations — install, onboarding, support × 5–8 sites$150K
Clinical validation — peer-reviewed study, CHAMP / collegiate$120K
DoD SBIR consultant + travel (AFWERX / SOFWERX)$60K
Fractional VP Sales — 3-month engagement, then extend$80K
Case studies, media, conference booths$40K

● IP

$300K
Utility patent conversion (3 claims)$45K
PCT filing — international protection$60K
Continuations (method + apparatus + system)$90K
Data-rights counsel — contract template, per-deal review$60K
Trademark, prior art, IP strategy$45K

● Reserve — legal, ops, contingency

$150K
D&O, product liability, general liability insurance$40K
Corporate legal, employment, cap-table management$35K
Accounting, audit, ops overhead$35K
Contingency — component lead times, rework, rush orders$40K
Total Raise $3,000,000

$3M seed gets Modulor through 5–8 pilot deployments, utility patent conversion, first insurance data-licensing LOI, and AFWERX Phase I submission → Series A at $2–4M ARR + corpus of 25K+ athlete-months.

14

The Ask

$3M seed — strategic-led. Three seats on the cap table.

The thesis is defined. The IP is filed. The product is scoped. We are raising the smallest round that buys 30 months and the right partners on the cap table.

Seat 1 · Anchor

Strategic Investor

One strategic who brings distribution or a data-licensing pathway. Hyperice, Therabody, Catapult, Nike, a major insurance carrier, or a sports-specific fund with operator relationships.

Check size:

  • $1–1.5M lead
  • Distribution or data-licensing intent letter
  • Board observer seat

Seat 2 · Partner

Sports-Specific Fund

A fund with explicit sports/health thesis and warm intros to pro + collegiate training rooms. Will lead or co-lead. Brings deal flow on future strategics.

Check size:

  • $500K–$1M
  • 3–5 warm intros to flagship pilot programs
  • Governance seat or observer

Seat 3 · Signal

Athlete / Operator Syndicate

Active and retired pro athletes + performance directors who take small positions and become public advocates. Signaling asset worth more than the check.

Check size:

  • $25K–$250K per participant
  • 3–5 named advocates at close
  • No board seat

What's done

  • USPTO provisional 64/040,028 filed Apr 15, 2026
  • Full product scope + BOM
  • Data-rights contract template
  • Pilot conversations active
  • Design docs public (19)

What $3M delivers

  • 5–8 pilot stations deployed
  • Utility patent + PCT
  • PNF at V1.5 — clinical parity
  • AFWERX Phase I submission
  • First insurance LOI

Milestones to Series A

  • $2–4M ARR
  • 25K+ athlete-months in corpus
  • 1 insurer data-licensing pilot
  • 3+ AMS integrations live
  • DoD non-dilutive awarded

If you are that anchor, that fund, or one of those athletes — let's talk.

Direct: hello@modulor.example · Data room: hub.html · Reply within 48 hours.

15

MODULOR

Controlled tension. Measured range. Data on every rep.

Confidential — 2026