During the toy development process, the project engineer is responsible for creating a comprehensive part list, essentially functioning as a Bill of Materials (BOM). This critical document serves as a foundation for the entire team, guiding activities such as prototyping, material sourcing, and tooling estimation.
However, the manual creation of part lists, especially for complex toys with numerous components, is time-consuming and prone to human error.
Manual Process: Part list creation is currently a manual process.
Error Prone: The manual process is highly susceptible to human error, such as typos.
High Costs: Errors in part lists can result in significant financial losses.
tools creation part list use IBM AS/400
database color#, registered part number such as clothes, plastic part and hair rooted, etc. also tooling information centralized to shared folder
create formatting part list in excel and user VBA to access share folder and creating part list in excel based on format we set
after that, send part list in excel to AS/400 by use VBA
Process creation part list improve by remove manual activities by ~80% of total
Increase productivity project engineer by able to finish task creating part list from 1-2 toys number in a week into 3-5 toys number
excel and VBA
OneDrive share folder
AS/400
Automated Color Code Extraction: Utilize Google AI or open-source AI to extract color codes directly from design sheets.
Streamlined Part List Generation: Integrate extracted color codes into an Excel-based part list format to expedite the creation process.
Prior to toy development approval, the project engineer convenes a DFMEA meeting with all relevant team members to conduct a thorough risk assessment. During this meeting, each team representative identifies potential concerns related to packaging, hairstyle, fashion, design, assembly, and coloring. The project engineer documents these concerns, analyzes their potential impact, and collaboratively develops preventive actions.
DFMEA meetings typically last 1-2 hours for standard toys and can extend to 3-4 hours for more complex products.
DFMEA meetings often conflict with team members' primary responsibilities, such as ongoing toy development or departmental improvements.
Analysis indicates that the current DFMEA process has not effectively prevented low lab test scores during development. These subpar results necessitate additional material resubmissions, leading to increased development costs.
Develop a centralized database to store existing DFMEA data, including identified concerns, potential effects, and preventive actions. This database will automate the review, input, and analysis processes, streamlining the DFMEA process.
When conducting a DFMEA, the team can compare the new toy to similar existing products. The system will suggest potential concerns, effects, and preventive actions based on these similarities, significantly reducing meeting time and allowing teams to focus on unique product characteristics.
Before each piloting stage, the quality team will require approval from the project engineer to confirm that the DFMEA process has been adequately addressed. This step aims to prevent potential failures by ensuring all necessary actions have been taken prior to lab testing.
First-pass yield for new toy submissions to lab increased from 70% to 93%.
This improvement resulted in a 12% reduction in potential defect costs compared to the previous year.
DFMEA meeting able to conduct virtually and improve duration meeting for complex toy from 3-4 hours into 1-1.5 hours
Excel
Catia
Exploring the integration of AI-powered 3D tools to automate DFMEA analysis through prompt-based exploration.
Currently, project engineers coordinate toy development through WhatsApp groups and shared Excel files. This decentralized approach has led to consistent challenges, including miscommunication and information overload due to multiple group chats.
Development meetings often yield vague updates like "will check later."
WhatsApp groups are an inefficient platform for project management
Teams require timely updates, especially during peak development periods
Develop a web-based platform, ONE PRODECO, to visualize project status using a Kanban board.
Integrate material traceability systems to provide real-time material updates.
Enable team members to add comments and track project progress.
Incorporate lab test results and rejection rates for data-driven decision making.
Eliminate WhatsApp groups and create a centralized communication hub.
Provide real-time project updates to internal and external stakeholders.
Communication in development only focus for preventive action or more valuable action instead "checking material etc" due to know all system information system in each department already integrated into one single website ONE PRODECO
Laravel
Visio
ONE PRODECO has the potential to become a best practice across Mattel's global operations. The SVP of Global Manufacturing and Operations Asia has requested its implementation in sister plants across China, Malaysia, and Thailand.
Project Engineers are tasked with locating existing parts, tools, and samples for designers in the US, Hong Kong, and Indonesia to facilitate toy concept development. This information is crucial for creating initial prototypes and presenting them to the marketing team before cost analysis.
The manual process of locating parts and samples is time-consuming and inefficient due to the lack of a centralized database. The physical storage of samples without an indexing system further complicates the process. As a result, finding necessary components for a single toy concept can take up to a week, impacting overall development timelines.
A part library within the ONE PRODECO system will be implemented to store detailed information about parts, including part number, category, storage location, mold details, and images. Integration with the AS/400 system will provide real-time mold status updates across global manufacturing sites.
The part library significantly reduced the time required to locate and prepare parts and data, accelerating the concept development process from an average of 14 days to 1-3 days. Global design teams have expressed enthusiasm for the improved efficiency and access to part information.
Laravel
Scan Barcode Tools
To further optimize the process, a request management system will be developed to track sample requests, improve accountability, and provide real-time status updates.
we create a Kaizen event in a month to improve all our job process into better. in this case we focus for improvement toy in development and production stage. we focus on how we reduce cost material or operational and make sure the aesthetic is not impact.
How we can reduce cost material or manhour but not impacting to aesthetic?
Conduct review based on structure cost, which element contribute as big cost
Evaluate whether any substitute or we can eliminate with constraint keep the aesthetic as same
Conduct mini piloting and submission to lab to make sure the improvement have same score / passing rate as original development
Conduct approval to Director Product Development and release the changes to production or development
In total a year, out team able to delivery 43% saving from original cost
Excel
Catia
as Mattel also one of 3500 companies participate to support sustainability. Mattel has target to achieve 100% use recycle part and packaging by 2030.
Mattel has no experience to use recycle material
How cost and aesthetic will impacted by this critical changes
at the beginning, Mattel Engineering Team focus on this. and Mattel Indonesia choose as trial place.
Learn character material from Material Data Sheet, makes comparison to existing character material we already use in production. this not only for plastic parts. but for all material we use to create a barbie doll
Create research sheet contain of comparison each material and how the result assembly, injection, etc. also result lab test per each process to make sure all process pass in every detail
Biweekly call with team Engineering, Quality, and Design to maintain Aesthetic, Quality, Cost as our target (team around the world from US, Hong Kong , and Indonesia)
Communicate all update and suspect things, so decision will make by together as team
we successfully launch our first doll with 80% of recycle material by 2021
this as good start to improve other 20% and we still have time until 2030
Injection mold
Assembly tools
Catia
PLC