Produced by Zhineng Auto
In this year's automotive market, the number of vehicle models is increasing, but the duration of popularity for a single model is shortening.
Li Bin used the proverb "No flower stays red for a hundred days" to describe the current industry situation. The hot-sales cycle for new vehicles has significantly shortened; a complete sales cycle might be as short as six months, making it difficult for any single model to maintain strong sales for a whole year.
In the early stage of a new product launch, demand concentrates and surges, leading to a rapid backlog of orders. However, after the enterprise struggles through capacity ramp-up, runs the supply chain at full load, and expands production lines with three-shift operations, market demand quickly falls back.
For OEMs (Original Equipment Manufacturers), the pace of launching new models, trial and error, and switching must be faster. This rhythm is quickly transmitted to the supply chain.
In an exchange with Cao Jin, Deputy General Manager of Xingyu Automotive Lighting, he used "five mores" to describe the current manufacturing environment:
◎ More Product Definitions
◎ Faster Delivery Pace
◎ Lower Cost Requirements
◎ Higher Quality Requirements
◎ Greater Order Uncertainty
For an automobile, a product that heavily emphasizes safety and quality, a core contradiction has emerged. Lean manufacturing focuses on producing things stably, yet now manufacturers must also adapt to changes at all times.
Li Bin's remark is very apt: "Just as capacity ramps up, orders disappear; just as the supply chain is fully equipped, demand drops. The severe mismatch in production and sales rhythms causes enormous waste across the entire industry chain." So, if this problem is posed to Dr. Cao, what would you do?
Part 1: Why Automotive Lighting Is Harder to Manufacture
Among all automotive components, automotive lighting is a relatively "unique" category.
Cao Jin mentioned that smart automotive lights are inherently multidisciplinary products, involving multiple fields such as mechanics, electronics, software, optics, and thermal management.
Every change can affect multiple links. This is compounded by several practical factors:
◎ High Degree of Customization with Significant Differences Across Vehicle Models
◎ Short Development Cycles Leaving Limited Time for Trial and Error
◎ Large Order Fluctuations Making It Difficult to Produce Entirely According to Plan
◎ Numerous New Features Without Any Compromise on Quality Requirements
Meeting numerous different requirements simultaneously within a short timeframe. In the face of such an environment, the most direct response from the manufacturing side is often to reduce costs, improve efficiency, and ensure quality.
These principles hold true at all times, but in this exchange, Cao Jin mentioned that if he had to choose the single most critical capability, he would lean towards "flexibility."
Why flexibility?
This issue can be viewed from a different perspective. If orders are stable and production is in large batches, efficiency is naturally the most important;
But if orders are constantly changing and products are continuously adjusted, then "switching capability" becomes more critical. For example:
◎ Whether a Production Line Can Quickly Switch Between Different Products
◎ Whether Equipment Utilization Can Remain Stable Under Multi-Variety Conditions
◎ Whether Small-Batch Orders Can Be Handled Without Significantly Driving Up Costs
Without flexibility, simply pursuing efficiency can instead easily lead to resource waste.
Over the past few years, automation has become a consensus in the manufacturing industry. At Xingyu, the density of industrial robots has currently reached a high level, and collaborative robots are also being applied on a large scale.
Automation solves the problem of stabilizing physical actions. The next step begins to involve "how to utilize data." Currently, factories have accumulated relatively complete data that can be used for operational monitoring.
Moving further forward, the goal is to use this data to assist in decision-making. For example:
◎ Which Links Are Prone to Issues
◎ Under Which Working Conditions the Yield Rate Is Higher
◎ When Equipment Requires Maintenance
This step is still in progress, but the direction is already quite clear.
Part 2: AI Is Already in Use in Some Areas, and robot application scenarios
The application of AI (Artificial Intelligence) in manufacturing has been widely discussed in recent years.
Within Xingyu, there are already several specific implemented scenarios:
◎ Document Inspection
◎ Assisted Code Development
◎ Equipment Operation and Maintenance Assistants
◎ Machine Vision Inspection (e.g., Scratch Recognition)
These applications share a common feature: they are mainly concentrated in highly repetitive tasks with relatively clear rules.
In these links, AI can indeed help reduce a significant amount of workload. However, there are still some areas where full automation is currently difficult. For instance, the plastic parts used in automotive lights are affected by material characteristics during positioning and assembly, still requiring a certain amount of debugging experience.
Such issues also involve empirical judgment, where AI serves more as an auxiliary tool rather than a complete replacement.
In this high-frequency changing environment, the development pace is also accelerating.
Cao Jin mentioned that Xingyu is advancing a clear direction: integrating styling, R&D, design, process engineering, electronics, mold making, and manufacturing as much as possible. Tasks that were originally scattered across different links are now consolidated into a single system for collaborative completion, as the cost of back-and-forth communication continues to rise.
When product changes are frequent:
◎ One Design Change Requires Corresponding Process Adjustments
◎ A Process Change Necessitates Synchronization on the Manufacturing Side
If these are scattered across different teams or even different companies, the cycle can easily be prolonged. Integrating these links as much as possible allows for faster adjustments and easier rhythm control. In terms of results, this helps both development efficiency and cost control.
On the manufacturing floor, robots have become very common. They are currently mainly applied in: screw locking, welding, glue dispensing, and handling, all of which are relatively standardized actions.
However, there are still some scenarios where single-arm robots are not very convenient to use, such as: wire harness insertion, multi-part synchronous assembly, and some complex inspections. These tasks are closer to manual operations and require more flexible combinations of movements.
Xingyu is also exploring the direction of multi-arm collaboration and even embodied robots.
Stability, takt time efficiency, and long-term operational reliability are several practical issues that are relatively difficult to solve.
Summary
The automotive manufacturing sector is undergoing a rather different phase.
In the past, it was more important to do one thing stably and at scale;
Now, there is an additional layer of requirement: maintaining stability amidst changes. Orders are changing, products are changing, and the rhythm is changing.
The manufacturing system can only adjust accordingly. Automotive lights are still the same automotive lights, but the manufacturing system of automotive lighting factories has undergone significant changes and is no longer the operating mode of the past.