EMS Capability Evaluation Insight

AMZ General VEM Level - Cut Off 2025/10/15

Product-Based Supplier Capability Overview

Capability Bar Chart

Displays the top 2 performing suppliers for each product category, ranked by their capability scores based on VEM level assessments across technical functions.

Scoring Formula (HC Savings):
ULE: 0.9 HC saved
L VEM: 0.8 HC saved
M VEM: 0.5 HC saved
H VEM: 0 HC saved

Scores represent Headcount (HC) savings. Higher scores indicate greater resource efficiency and reduced engagement requirements.

Supplier Legend

Hover over legend items to highlight corresponding bars in the chart

Each bar shows supplier name and score on the right side

Click on product names to jump to Product Function Analysis

HQ DB1
HQ DB2
GTK
CNT
FIH
GGEC

Product Function Analysis

Select a product to view each supplier's capability scores across different technical functions. The radar charts show VEM level scores together with average scores.

Supplier Score Comparison: 2024 vs 2025

Interactive Supplier Selection

Click on a supplier below to view their 2024 vs 2025 score comparison

HQ DB1
HQ DB2
GTK
CNT
FIH
GGEC
Chart Information:

Shows 2024 (dashed line) vs 2025 (solid line) scores for the selected supplier across 8 technical domains.
Scores are based on the overview page scoring rules: ULE(100), L VEM(80), M VEM(60), H VEM(20)

Supplier Technical Capability Assessment Matrix

Category Product HQ DB2 HQ DB1 GTK CNT FIH GGEC
Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra Audio HW Audio SW Camera DOT Sensor Wireless Production Test SW Infra

HQ-DB1 (Category: Hearable)

High Risk Level

Evaluation Result:

HQ-DB1 does not qualify as an ULE partner for four critical reasons:

  1. Sensor testing fixtures are mostly manual operating, bench automation and fixture spec can be observed as a gap to Amazon standard
  2. New sensors like mid-IR proximity sensor with skin detection function and BIO sensors need further study to support
  3. Lack of wireless systematic risk evaluation
  4. No experience to do correlation on ARF-OTA performance for running effectively FA on antenna/ARF issue during NPI

ULE Mitigation Plan:

ULE Design Playbook
  • Establish a playbook mechanism and conduct daily design review sessions with key stakeholders before EVT RR to ensure all technical requirements and specifications are met.

Suggestion:

JDM (M VEM) before EVT RR

HQ-DB2 (Category: General)

Mid Risk Level

Evaluation Result:

HQ-DB2 is an ULE partner for Echo, eReader, and Tablet products. However, they must provide a robust solution to ensure compliance with BRD standards during their transition to HW Specifications. Additionally, HQ-DB2 needs to implement preventive measures to avoid similar compliance issues in future projects.

ULE Mitigation Plan:

ULE Design Playbook
  • Amazon connectivity design team can develop a ULE playbook and scheduled design reviews before the EVT RR.

Suggestion:

JDM (L VEM) before EVT RR

GTK (Category: General)

High Risk Level

Evaluation Result:

GTK does not qualify as an ULE partner for six critical reasons:

  1. Camera does not have IQ advanced feature tuning, subjective & objective tuning, and AIISP tuning methodand
  2. Current labs are not sufficient to support MTK/Qualcomm ISP tuning, no DNP light source, and the camera labs are not qualified by Amazon yet
  3. Sensor is missing Accel test bench, P-Sensor Test bench, and TOF Test bench Develop
  4. GTK wireless design and production capability could not offload NTI loading, such as radar, A2S, and simulation requirements
  5. No experience on Complex issue Troubleshooting Capability improvement-De-sense
  6. No DOT qualified Sr. engineers and No pDOT design, testing, and validation experience

ULE Mitigation Plan:

ULE Design Playbook
  • Establish a standardized Hardware (HW) specification mechanism and begin implementation with the JDM to ensure alignment on technical requirements and quality standards.
  • For Eacho Core and Entry Vnext, we developed a ULE playbook and scheduled design reviews before the EVT RR

Suggestion:

JDM (M VEM) before EVT RR

CNT (Category: General)

Mid Risk Level

Evaluation Result:

CNT, with over 10 years of experience working with Amazon on OEM business models, specializes in SMP-TV Stick projects. To support CNT's transition to an independent ULE model for the SMP-TV Stick, Amazon connectivity design team has developed a ULE playbook and scheduled design reviews before the EVT RR. Based on this preparation, CNT has been awarded the Mercer project (next generation TV stick) under the ULE model.

ULE Mitigation Plan:

ULE Design Playbook
  • Amazon connectivity design team can develop a ULE playbook and scheduled design reviews before the EVT RR.

Suggestion:

JDM (L VEM) before EVT RR

FIH (Category: General)

High Risk Level

Evaluation Result:

FIH-General does not qualify as an ULE partner for three critical reasons:

  1. No qualified team involvement in Brioche project design
  2. Laboratory infrastructure not established prior to Brioche engagement
  3. Recently hired senior engineers lack Amazon DRIs validation

ULE Mitigation Plan:

ULE Design Playbook
1. Validation Level Adjustment
  • Increase Amazon VEM requirement to Mid/High
  • Leverage Amazon R&D resources for schedule adherence
2. Laboratory Resource Optimization
  • Utilize existing Amazon laboratory facilities
  • Require EMS to reduce device MVA
  • Eliminate delays from EMS lab setup
3. Engineering Qualification Process
  • Verify device simulation expertise of EMS senior engineers
  • Complete Amazon DRI qualification before RFQ release

OEM-to-ULE Conversion Plan:

By December 31, 2025:
  • EMS laboratory must complete setup
  • All Sr. engineers must secure capability validation from Amazon DRIs
By February 26, 2026:
  • Obtain Amazon laboratory certification

Suggestion:

OEM

FIH (Category: Seabright and Stinson)

High Risk Level

Evaluation Result:

FIH-Seabright and Stinson do not qualify as an ULE partner for four critical reasons:

  1. High engagement for camera since it has V0 features (night mode)
  2. Sensor needs re-architect RGB sensor (inferencing with camera night mode IR LED as well) and supports omnisense
  3. Wireless has new WiFi radar feature (2T|1R) and use new WiFi Chip IC 7903
  4. No ability to develop Production HW (Chamber) and SW (WFT)

ULE Mitigation Plan:

Hardware specification mechanism
  • Establish a standardized Hardware (HW) specification mechanism and begin implementation with the JDM to ensure alignment on technical requirements and quality standards.

Suggestion:

OEM

GGEC (Category: Hearable)

High Risk Level

Evaluation Result:

GGEC does not qualify as an ULE partner for eight critical reasons:

  1. Sensor has gap in function characterization, device bench testing fixtures, manual or automatic
  2. GGEC will need AMZ team's support to build sensor lab and related capability
  3. New sensors like mid-IR proximity sensor with skin detection function and BIO sensors need further study to support
  4. Lack of wireless systematic risk evaluation
  5. Minimal SAR testing experience and Not familiar with FCC/CE SAR test waiver conditions for TWS
  6. No expereience to do corerlation on ARF-OTA performance for running effectively FA on antenna/ARF issue during NPI
  7. No dedicated sensor validation capabilities, desense validation(Spectrum capture), and cross body effect validation
  8. Not familiar with Amazon production testing requirement

ULE Mitigation Plan:

Hardware specification mechanism
  • Establish a standardized Hardware (HW) specification mechanism and begin implementation with the JDM to ensure alignment on technical requirements and quality standards.

Suggestion:

JDM (M VEM)

Technical Domain Detailed Information

Audio SW Domain

Audit Date: HQ DB1 10/10/2025, GGEC 10/14/2025 Focus: Capability in hearable audio

HQ DB1 Assessment

Score: 2.91/5
✓ Strengths
  • Anechoic chamber owned by the BU, can share labs from other BUs
  • All required equipment available (HATS, ACQUA, 3QUEST, Soundcheck, AudioPrecision)
  • Experience in all kinds of earbuds and wearable products
  • Experience in tuning ANC, EQ algorithms
  • Experience in major earbuds solutions
  • Following same procedure as other BUs working with Amazon
⚠ Weaknesses
  • No ETSI room nor Listening room
  • Lack of ASR experiences
  • Limited experience in ASR (Subjective only) and VoIP

GGEC Assessment

Score: 2.95/5
✓ Strengths
  • Certified chamber following multiple international standards
  • Extensive experiences in ANC, ENC tuning and algorithm design
  • In-house playback algorithm (GSound)
  • Strong tuning capability in adaptive ANC ENC and EQ algorithms
  • Cross productline software stack experience
⚠ Weaknesses
  • No ETSI room nor Listening room
  • No ACQUA nor 3QUEST equipment
  • No more than two dedicated audio software engineers
  • Lack of ASR VoIP experiences
  • Does not follow well defined PDP procedure in software dev/release

Sensor Domain

Update Date: 2025/9/25
📊 General Assessment
  • Wearable products are High engagement for all EMS due to complexity
  • HQ: Tablet is already ULE in sensor domain
  • GTK: Turn eReader and Tablet to Mid as they have never touched those models
  • CNT: Lacks sensor characterization equipment and Amazon sensor program experience
  • FIH: Tablet is Mid as they have never been awarded a tablet program

HQ DB1 Assessment

2025-10-10
✓ Capabilities
  • Familiar with common sensors in TWS products (hall, capacitive, IR proximity, IMU, resistive pressure)
  • EE design team capable of chip selection, architecture design, sensor system level simulation
  • Experience in sensor SMT/FATP testing support, calibration, tuning, FA and device bench testing
  • Lab fixtures and process cover sensor testing requirements and OBA testing
⚠ Gaps
  • Gap in algorithm development and evaluation
  • Sensor testing fixtures mostly manual operating
  • Bench automation and fixture spec gaps to Amazon standard
  • New sensors (mid-IR proximity with skin detection, BIO sensors) need further study

GGEC Assessment

2025-10-14
✓ Capabilities
  • EE HW design team familiar with common TWS sensors
  • Capable of chip selection, architecture design, sensor system level simulation
  • Demonstrated in-house and 3rd party algorithm development capability
⚠ Gaps
  • Gap in sensor function characterization and device bench testing fixtures
  • Need Amazon team support to build sensor lab and related capability
  • New sensors (mid-IR proximity with skin detection, BIO sensors) need further study
💡 Remark

VEM depends on system requirements case by case. Complex sensor stacks may change VEM from Low to Mid/High.

Wireless Domain

HQ DB1 Assessment

🚨 Antenna Design Risks
  1. Antenna design analysis and simulation falling short of Amazon expectations
  2. No experience evaluating worst case scenarios (human cross body impact, component variations)
  3. Only used CST default head model, not aligned with P10 model standard
  4. No human cross body simulation and validation capability
  5. No CTIA standard antenna OTA chamber in factory/lab (sharing Mini/Micro chambers with BD2)
  6. No significant experience in ARF-OTA performance correlation for effective FA during NPI
⚠ RF System & Desense Issues
  1. HQ DB1 team not familiar with Amazon RF&Desense design PDP
  2. Desense simulation capability rarely used for risk evaluation
  3. Desense risk assessment relies on experience, lacks systematic evaluation
  4. No consideration of antenna detune risk in ARF station factory testing

GGEC Assessment

🚨 Antenna Design Risks Very High Risk
  1. Lacks experience in worst-case scenario evaluation
  2. P10 head model parameters not aligned with standard (using default CST settings)
  3. No human cross-body simulation and validation capabilities
  4. Limited ARF spec window setting experience and antenna performance correlation
  5. Minimal SAR testing experience, unfamiliar with FCC/CE SAR test waiver conditions for TWS
  6. Insufficient ARF-OTA performance correlation experience for NPI FA
  7. Unable to develop solid high-level validation schedule from Amazon wireless HW specs
  8. Only one MVG SG24 OTA Chamber in R&D lab, lacks factory-based OTA chambers
  9. Insufficient knowledge of 3D-printed mockup materials (SLA, PLA, PTEG, PC, ABS)
⚠ RF System & Desense Issues
  1. Poor CSE/RSE design and failure analysis capability
  2. No ZigBee signaling equipment and power table experience
  3. Poor desense validation capability, not following Amazon standards
  4. Poor coex design capability (best performance: 25dB isolation between 2.4GHz and ZigBee)
  5. Poor systematic desense design capability

Validation Domain

GGEC Assessment

Overall Risk: Medium to High
📡 Wireless Validation Medium Risk
✓ Capabilities:

Basic validation facilities (OTA chamber, shielding room, required instruments)

📋 Test Plans:

OTA, antenna performance, conducted testing for RF basic performance and pre-compliance

⚠ Issues:

Test limits problematic (conducted spurious emission limit only -40dBm/MHz, likely causing radiated compliance failure)

❌ Gaps:

No dedicated desense test station, no validation plan for wearable devices, no cross body analysis capability

🔍 Sensor Validation High Risk
  • No dedicated sensor validation capabilities
  • No dedicated sensor engineer for validation
  • No validation setup or plan for various sensors
🎵 Audio Validation Low Risk
  • GGEC has almost all required equipment for HW validation
  • Refer to Audio SW assessment for detailed capabilities
⚡ Other EE Validation
  • Available: Basic EE validation facilities
  • Capabilities: PI/PDN/SI/Ripple simulation and measurement
  • Need evaluation: Sophisticated layout and PCB stackup planning integration capability

Production Test Domain

Assessment Date: 2025/10/10 (HQ), 2025/10/14 (GGEC)
📊 Supplier Scores Summary
Supplier Score Engagement Level HC Needed
HQ 88 0.2
GTK 88 0.2
CNT 88 0.2
Ensky (FIH) 88 0.2
GGEC 81 0.5

HQ-DB1 Detailed Assessment

Region: Nanchang Score: 88/100 Evaluator: hanry@amazon.com
✓ Strengths
  • Experience in comprehensive device testing for large-scale manufacturing
  • Proficient in implementing Presto testing systems in production environments
  • Expertise in standardized quality control procedures
  • Efficient engineering resource management for testing operations
  • Perfect scores in Equipment, Human Resource, Maintenance, Automation, and Failure Analysis (100%)
⚠ Weaknesses
  • Limited ARF chamber capabilities prevent comprehensive design validation during NPI
  • Experience category scored only 52% due to fixture design capability gaps
  • Need collaboration with Lab126 Production Engineering team for holder design optimization

GGEC Detailed Assessment

Region: GuangZhou Score: 81/100 Evaluator: hanry@amazon.com
✓ Strengths
  • Experience in comprehensive device testing for large-scale manufacturing
  • Proficient in operating RF equipment for troubleshooting and annual calibration
  • Expertise in standardized quality control procedures
  • Perfect scores in Equipment, Human Resource, Maintenance, Automation, and Failure Analysis (100%)
⚠ Weaknesses
  • Lack of ARF chamber design capabilities prevent comprehensive design validation during NPI
  • Experience category scored only 36% due to significant fixture design capability gaps
  • Need collaboration with Lab126 Production Engineering team for holder design and antenna placement optimization
📋 Score Definition
  • 95-100: Ultra Low Engagement (ULE)
  • 86-95: Low Engagement (LE)
  • 76-85: Mid Engagement (ME)
  • 61-75: High Engagement (HE)
  • <60: Disqualified

SW Infra Domain

HQ NC DB1 Assessment

2025-10-10
🖥️ RF Test Software Platform
Platform: I-Far testing platform supporting all test stations
Chipset Support: Multiple chipsets including MTK, Qualcomm, and other TWS chipsets
✓ Capabilities
  • Comprehensive test station support through I-Far platform
  • Multi-chipset compatibility for various TWS solutions
🚨 Critical Gaps
  • Lacks RF signaling test software development experience
  • I-Far test platform only supports Non-Signaling tests
  • Presto RF test instrument not included in current test platform (needs upgrade)
  • I-Far platform does not support BT Tx calibration (needs upgrade)

GGEC Assessment

2025-10-15
⚠ Resource Limitations
  • Lacks resources to support factory line test tool development
  • Only one software engineer supports all line test stations' tools and MES integration
❌ Development Capability Gaps
  • Lacks RF and Audio test tool development experience
  • 90% of production line test tools outsourced to third-party vendors
  • Heavy reliance on external vendors for critical testing infrastructure
📈 Impact Assessment
  • Limited internal development capability may impact customization and rapid response to testing requirements
  • Single point of failure with only one dedicated software engineer
  • Vendor dependency may affect cost control and timeline flexibility