Low-Noise Gearbox Design for Hospital and Retail AMR: Acoustic Engineering Guide
Low-noise gearbox engineering for hospital and retail AMR. Noise thresholds by environment, architecture comparison, and test protocols.
In hospitals and retail sites, drivetrain noise is a deployment risk. If the robot is perceived as intrusive, program expansion slows regardless of task performance. Many teams discover this too late because they review only overall dBA at one speed, then miss tonal peaks that trigger complaints.
Noise Thresholds by Deployment Environment
Before selecting a gearbox, understand the ambient noise baseline and acceptable robot contribution:
| Environment | Ambient noise | Max robot target | Critical frequency | Complaint trigger |
|---|---|---|---|---|
| Hospital (night ward) | 30–35 dB(A) | ≤ 40 dB(A) @ 1m | 500–2000 Hz | Tonal whine during sleep hours |
| Hospital (corridor) | 45–55 dB(A) | ≤ 50 dB(A) @ 1m | 1000–4000 Hz | High-pitched gear mesh |
| Retail (open floor) | 55–65 dB(A) | ≤ 55 dB(A) @ 1m | 2000–6000 Hz | Acceleration whine near customers |
| Hotel / library | 35–45 dB(A) | ≤ 42 dB(A) @ 1m | 500–3000 Hz | Any audible tonal component |
| Office / lab | 40–50 dB(A) | ≤ 48 dB(A) @ 1m | 1000–4000 Hz | Persistent whine during work |
| Warehouse / factory | 65–85 dB(A) | ≤ 70 dB(A) @ 1m | N/A | Usually not a constraint |
Key insight: A robot at 48 dB(A) may be perfectly acceptable in a warehouse but generate complaints in a hospital corridor. Always define noise targets relative to deployment environment.
Gearbox Architecture Noise Comparison
Trade-off: Lower-noise architectures (harmonic, worm) typically sacrifice efficiency. A worm gearbox at 48 dB may consume 15–20% more battery than a planetary at 58 dB. The selection must balance acoustics with energy budget.
Noise Sources in a Gearbox Drivetrain
Understanding where noise originates helps target the right countermeasures:
| Noise source | Frequency range | Primary gearbox types affected | Mitigation strategy |
|---|---|---|---|
| Gear mesh excitation | 500–6000 Hz | Spur, planetary, bevel | Micro-geometry optimization, profile crowning |
| Bearing noise | 1000–8000 Hz | All types | Preload optimization, quality grade selection |
| Housing resonance | 200–2000 Hz | All types (thin wall) | Rib stiffening, damping compound |
| Motor cogging torque | 100–500 Hz | All (motor-dependent) | Skewed magnets, sinusoidal commutation |
| PWM switching noise | 8000–20000 Hz | All (controller-dependent) | Higher switching frequency, LC filtering |
| Lubricant churning | Broadband | High-speed planetary, spur | Optimized fill level, viscosity selection |
Tonal vs Broadband Noise — Why dB(A) Alone Misleads
Noise Reduction Design Levers
| Design lever | Noise reduction potential | Trade-off cost |
|---|---|---|
| Gear micro-geometry optimization | −3 to −8 dB | +15–25% gear manufacturing cost |
| Helical instead of spur teeth | −5 to −10 dB | Axial thrust bearing required |
| Bearing preload optimization | −2 to −5 dB | May reduce service life if over-preloaded |
| Housing rib stiffening | −3 to −6 dB | +5–10% weight, +10% casting cost |
| Vibration isolation mount | −5 to −10 dB | May reduce positional accuracy |
| Lubricant viscosity optimization | −1 to −3 dB | Temperature range narrowing |
| Motor sinusoidal commutation | −3 to −6 dB (cogging) | Controller cost increase |
| Higher PWM frequency (>20kHz) | Eliminates audible switching | Increased switching losses |
Production-Intent Test Protocol
Low-noise claims from bench fixtures often fail after integration. Require tests with production-intent stack:
Minimum test matrix
| Condition | Speed points | Load points | Temperature | Samples |
|---|---|---|---|---|
| Cold start | 3 speeds | No load + 50% load | Ambient | ≥ 3 units |
| Warm state (stabilized) | 3 speeds | No load + 50% + 100% | Stabilized | ≥ 3 units |
| Acceleration sweep | Ramp 0→max | 50% load | Both | ≥ 2 units |
| Direction reversal | 3 speeds | 50% load | Warm | ≥ 2 units |
Data deliverables per test point
- Overall dB(A) with measurement distance and microphone position
- 1/3 octave band spectrum (minimum) or narrow-band FFT (preferred)
- Prominent tone identification with frequency and level
- Background noise level during test
Acoustic Quality Gates
| Gate | Timing | Pass criteria |
|---|---|---|
| G1 — Concept | Before RFQ | Noise target defined by environment, spectrum requirements specified |
| G2 — Sample | Engineering sample | dB(A) and spectrum meet target at bench, no disqualifying tones |
| G3 — Integration | Full robot build | System-level noise meets target in production-intent config |
| G4 — Pilot | Field pilot | No noise complaints from operators or bystanders over 2-week trial |
| G5 — SOP | Production lot | Lot-level sampling within ±2 dB of G3 baseline |
RFQ Noise Specification Template
Add these mandatory fields:
| Field | Requirement |
|---|---|
| Noise target | Overall dB(A) at [speed, load, distance, temperature] |
| Spectrum format | 1/3 octave or narrow-band FFT, frequency range 100 Hz–10 kHz |
| Prominent tone limit | No tone exceeding background + 6 dB in 500–4000 Hz band |
| Test setup | Production-intent assembly, defined fixture, calibrated microphone |
| Warm-state requirement | Minimum 30-minute thermal stabilization at test load |
| Sample count | Minimum 3 units from different production lots |
| Measurement report format | Include background noise, setup photo, raw data |
Buyer Decision Rule
For hospital and retail AMR programs, choose candidates that deliver consistent acoustic behavior in integrated conditions, not just the best single-point lab number. Predictability across temperature, speed, and assembly variation is the main risk reducer.
Priority ranking for noise-sensitive applications:
- Spectrum quality (no prominent tones) > Overall dB(A) level
- Warm-state performance > Cold-start performance
- System-level result > Bare gearbox result
- Multi-sample consistency > Single-unit best case
Related Engineering Guides
- Planetary vs Cycloidal vs Harmonic — Noise levels compared across gearbox types
- Compact Gearbox for Sub-300mm AMR — Packaging impacts on acoustic behavior
- OEM Customization Checklist — Specify noise targets in your customization request
- Browse Harmonic Drive Products
For low-noise gearbox screening and OEM customization support, contact [email protected].
Frequently Asked Questions
What noise level is acceptable for hospital AMR robots?
Hospital AMR robots should produce no more than 40 dB(A) at 1 meter during night ward operation and 50 dB(A) in corridors. More importantly, the robot must avoid tonal peaks in the 500–2000 Hz range, which trigger sleep disturbance complaints even at lower overall levels.
Which gearbox type is quietest for indoor AMR applications?
Harmonic drives are typically the quietest at 42–52 dB(A), followed by worm gearboxes at 45–55 dB(A). However, worm drives sacrifice 15–20% efficiency. For balanced noise-efficiency performance, well-optimized planetary gearboxes with helical gears can achieve 50–58 dB(A).
Why do two gearboxes with the same dB(A) rating sound different?
Overall dB(A) is a weighted average that masks tonal characteristics. A gearbox with smooth broadband noise at 52 dB(A) may be unnoticeable, while another at 52 dB(A) with a prominent 1.8 kHz tonal peak will generate complaints. Always request frequency spectrum data, not just overall dB(A).
How can I reduce gearbox noise without changing the gearbox type?
Key noise reduction levers include: gear micro-geometry optimization (−3 to −8 dB), vibration isolation mounting (−5 to −10 dB), housing rib stiffening (−3 to −6 dB), bearing preload optimization (−2 to −5 dB), and switching to sinusoidal motor commutation (−3 to −6 dB for cogging noise).
Author
Categories
Need an AMR Gearbox Recommendation?
Reference this article and share your constraints. Our team replies via [email protected].
More Posts
AMR Gearbox RFQ Template: Copy-Ready Structure for Technical Evaluation
Copy-ready AMR gearbox RFQ template with scoring model, compliance matrix, and red-flag questions for faster supplier evaluation.
How Gearbox Efficiency Impacts AMR Battery Life: Quantitative Analysis
How gearbox efficiency impacts AMR battery life. Quantitative energy loss model with fleet OPEX calculations and worked examples.
BLDC Motor + Planetary Gearbox Sizing for AMR Wheel Drives
Step-by-step BLDC motor and planetary gearbox sizing for AMR wheel drives. Includes formulas, worked examples, and thermal analysis.