A smart home company with an annual revenue of 120 million, with a R&D team of 46 people, and a product line spanning three major categories: smart door locks, smart lighting, and smart security.
During the investigation, project manager Wang Gong opened the “shared disk” displayed on his computer and found hundreds of Excel sheets, covering BOM lists, material codes, drawing version records, change application forms, etc.
However, an incident in which the wrong PCB board version number was used in a mass production batch of smart door locks made the company realize the seriousness of the management problem.
Before the transformation, enterprises had many core pain points.
BOM management relies on “person-to-person”. Different engineers maintain the BOM in Excel independently. There are no unified rules for material coding. Integrating the BOM of the entire machine requires manual comparison for 2-3 days.
The drawing version is like a “Russian matryoshka doll”. The shared disk folder names are confusing and it is difficult to find drawings. As the number of products increases, cross-referencing of drawings is prone to errors.
Change management relies on “verbal communication”. About 30% of R&D changes have no formal process, making after-sales traceability difficult.
Cross-department collaboration relies on “asking”, and information asymmetry leads to a lot of ineffective communication.
Supplier collaboration relies on “dumping drawings by email”, and the problem of mixing old and new drawings frequently occurs.
This company did not start with ERP like many manufacturing companies, but chose to solve the problem of R&D source data management first.
They believe that ERP handles “predetermined” data, and their own BOM is confusing. Going to ERP first will lead to data errors.
So we decided to implement a PLM system for “source management” first.
During the selection process, the company inspected three manufacturers.
Vendor A has full functions but high price and long implementation cycle; Vendor B has rich industry experience but heavy product architecture and complicated customization; in the end, it chose the third-grade PLM system because of its lightweight deployment, functional modularity, and support for rapid go-live.
Core considerations include speed of getting started, CAD integration capabilities, closed loop of change, as well as price and flexibility.
The implementation is divided into three stages.
In the first stage, we laid the foundation and spent nearly 6 weeks unifying data standards, including coding cleaning, BOM reconstruction and drawing archiving. This stage was very grueling.
The second phase of the process was launched, and the three core processes of ECR/ECO change management, BOM approval and release, and document version control were launched. However, resistance was encountered in the early stages of implementation, and strategies such as retaining the “quick change channel” were adopted.
In the third stage, all employees are promoted to enforce and optimize the process through administrative orders, and at the same time publicize early success cases, so that employees can change from “must use” to “willing to use”.
Half a year after its launch, the results have been remarkable.
The BOM arrangement time for a single product was reduced from 2 to 3 days to 0.5 days, and the efficiency was increased by 300% +; the drawing search time was reduced from 15 to 30 minutes to within 1 minute, and the efficiency was increased by 15 times +; the change closed-loop cycle was shortened by 60% +; material procurement errors caused by design changes were reduced to 0; the BOM accuracy rate was increased by 27%; the cross-department information confirmation time was reduced by 75%; the number of supplier collaboration errors was reduced to 0; and the completeness of R&D documents was increased by 63%.
There are three key factors for successful enterprise transformation.
The leadership has “real” support, and the R&D director personally handles projects to make employees aware of the importance of transformation.
Solve the biggest pain points first, focus on BOM management, change traceability, and drawing versions, and gradually expand.
Respect the habits of engineers and retain the Excel-style BOM editing experience in PLM.
For companies preparing to launch a PLM system, it is recommended not to save time on data cleaning, but to select “medium difficulty” for pilot products and bring in suppliers.
It took this smart home company four and a half months to complete the transformation from shared disk chaos to a controllable, traceable, and collaborative R&D management platform, proving that digital transformation is not a patent of big manufacturers. The key is to find a rhythm and method that suits you.
