On November 29th, ENOCH unveiled its "Technological Innovation Driving the Future" Automechanika Shanghai 2023 exhibition. ENOCH not only systematically exhibited its "Digital intelligence high-end quality refinishing total solution", "ENOCH intelligent color mixing system" and "full-stack digitalized vehicle after-sales service platform", but also officially released its ENOCH SPECTROPHOTOMETER, which attracted widespread attention from industry insiders. The release of the ENOCH SPECTROPHOTOMETER will further solve the industry's difficulties in paint color mixing, large color difference, and low efficiency, and lead the sheet metal and painting industry into the era of intelligent color mixing.

How did ENOCH establish its position and break the impasse in the long-standing problem of painting color matching?

In the whole process of sheet metal spraying, color mixing is undoubtedly the most critical part. The inefficiency of color mixing often leads to the inefficiency of the whole sheet metal spraying business. Traditional color matching relies heavily on experience, often requiring multiple manual plate comparisons, which not only has high time costs, large color differences, and weak controllability, but also relies heavily on talent. If we say that ENOCH's high-end quality spraying solution has broken through the traditional color mixing bottleneck through 65W+ real vehicle formula data and standardized IoT color mixing process, then ENOCH SPECTROPHOTOMETER has made color mixing truly free from "empiricism" and taken an important step towards the intelligent color mixing transition in the sheet metal and painting industry.

Difficulties and breakthroughs: Uncovering the challenges of car paint color matching

Although car painting is common in the process of car maintenance, it is not easy to restore the 1:1 paint color. Intelligent color measurement and matching solution refers to accurately interpreting the complex spectrum and particle information of the original car paint without relying on visual recognition, and providing quantitative color formula to perfectly replicate the paint information, presenting a color effect almost identical to the original car. However, both accurate color measurement and intelligent color matching have complex underlying logic.

The complex composition of car paint

Take the repair paint as an example, the thickness of the paint surface is about 115-165um (1mm=1000um), but the amount of information it contains is quite large, and its complexity mainly comes from the spatial hierarchy of the paint surface and the type of paint effect.

From a spatial perspective, the car refinishing paint consists of a bare metal layer, an epoxy primer layer (10-15um), a leveling layer (Sheet metal ash), a midcoat primer layer (50-70um), a color coat layer (15-20um), and a clear coat layer (40-60um), which are stacked from bottom to top. The complex light transmission of reflection, refraction, scattering, diffraction, and other light transmissions caused by the multi-level stacking greatly increases the difficulty of calculating the color mixing model of car paint.

From the perspective of paint effect, there are three common effects of car paint: plain paint, metallic paint, and pearl paint. The color of plain paint comes directly from the paint surface and does not change greatly with observation angle, so the color mixing is relatively simple. However, in reality, a large number of car paints need to add particle effect color masterbatch such as silver powder or pearl, which adds complex micro-surface texture characteristics and exponentially increases the amount of paint color data.

Difficulties in algorithm development

Due to the complexity of the surface information of car paint, accurate color matching requires more advanced technology and algorithms. The complex optical properties of particle effect color masterbatch lead to unstable physical properties, which are easily affected by external factors such as spraying process, environmental temperature and humidity, and drying time. The data changes are extremely sensitive, which poses great challenges to the determination of the relationship between coating formulation and the visual presentation of paint color and texture.

Firstly, the particle size of the granular color masterbatch is extremely small, reaching the micrometer (um) level. Therefore, advanced image processing algorithms such as deep learning models are required for high-precision processing to improve the accuracy of analyzing small particle color masterbatches.

Secondly, the color of the particle masterbatch exhibits angular dependence. Under different lighting conditions, the visual effects of the paint vary depending on the complex phenomena of specular reflection, diffuse reflection, scattering, and interference of light on the paint surface. The angular dependence introduces ambiguity in the data, requiring algorithms to accurately interpret this color variation. This requires further integration with complex optical models, such as BRDF (Bi-directional Reflectance Distribution Function) and BSDF (Bi-directional Scattering Distribution Function), to simulate and understand the changes in color under different angles.

Thirdly, the particle color masterbatch will be added with "color contamination" by the plain color masterbatch, which increases the complexity and impurity of the color data. Therefore, advanced spectral analysis techniques combined with spectral models are needed to distinguish and accurately process the interactions between different color masterbatches.

Finally, the paint color masterbatch does not have additive properties, which makes it difficult to quantify the color mixing equation in the formula and establish an accurate color mixing model. Traditional algorithms based on the color mixing theory cannot accurately solve the color matching problem. Therefore, it is necessary to use deep learning algorithms to analyze the characteristics of the color masterbatch and color mixing equation based on the real vehicle formula database, in order to more accurately solve the problems of paint color matching and toning, and make the information processing of the paint surface more efficient.

From 0 to 1 independent research and development, ENOCH innovative intelligent color matching solution

Multi-angle and multi-dimensional data analysis and determination

At the microscopic level, the many characteristics of particle color masterbatch bring unprecedented challenges to improving the accuracy of color data measurement. The ENOCH SPECTROPHOTOMETER is equipped with multi-angle directional illumination and a spectral analyzer with a pickup channel, with three measurement angles of 15°, 45°, and 110°, simulating the human eye's observation method. Within the visible wavelength range of 400-700nm, it collects one spectral data every 10nm, with a total of 31 sampling points to simulate a complete spectrum, to achieve high-resolution acquisition of paint spectral values, breaking free from the constraints of unobservable microscopic information of paint.

The difficulty in distinguishing particles after mixing has always been a technical problem that traditional color mixing cannot overcome. The intelligent ENOCH SPECTROPHOTOMETER uses high-precision cameras at four angles (80°, 45°, 15°, -15°) to accurately distinguish key values such as particle thickness, scintillation, and color scintillation that affect color presentation, fits out fine HDR image information of the paint surface, fully restores the complex micro-surface texture characteristics of the paint surface, and finally achieves accurate measurement of the microscopic characteristics of the paint surface.

ENOCH SPECTROPHOTOMETER solves the problems of traditional color measurement, such as difficulty, large color difference, and inaccurate identification of particle color masterbatch, by conducting multi-angle and multi-dimensional data analysis and measurement on paint surfaces, achieving fast, accurate, and efficient measurement of vehicle paint color information.

650,000+ real car color formula database support

After solving the color measurement problem, another major obstacle faced ENoch. For the development of an intelligent formula system, a large amount of real-car data needs to be collected. Without a certain amount of data accumulation, there is no data foundation for algorithm development, and the accuracy of the intelligent color algorithm cannot be guaranteed.

However, after seven years of practical accumulation, the color formula database of ENOCH has already accumulated more than 650,000 pieces of data, which means that most models on the market, regardless of production batch, production year, and color number, can almost find corresponding color matching schemes in the ENOCH database. Moreover, ENOCH covers more than 530 outlets in 28 provinces, municipalities, and autonomous regions, 160 cities across the country, supports real-time sharing and uploading of vehicle data formulas, and continues to increase at a daily growth rate of more than 500 pieces. These color matching data will be output to the intelligent color algorithm based on big data, supplemented by artificial intelligence training, to continuously optimize data quality, thus achieving a significant improvement in paint mixing efficiency.

Four original intelligent color algorithms, visual color mixing, revolutionize the efficiency of paint mixing

In the field of intelligent color matching, there is currently a lack of relevant theoretical guidance in the market, especially the lack of systematic paint color matching algorithm theory to draw on, which poses great challenges for R&D teams. The progressiveness of ENOCH SPECTROPHOTOMETER is that it carries four original intelligent color algorithms, including "intelligent matching algorithm", "intelligent color matching algorithm", "intelligent generation algorithm" and "car paint rendering algorithm", which inject new vitality into intelligent color matching.

Firstly, the "Intelligent Matching Algorithm" is based on the output of the multi-dimensional measurement data from the ENOCH SPECTROPHOTOMETER, constructing a multi-modal deep learning model algorithm to accurately match the output of the formula using spectral data, image data, and texture feature data. Relying on the ENOCH 650,000+ real vehicle color formula database, the material properties, optical properties, image properties, process properties, etc. are quantified, and the data collected from the paint surface are characterized and supplemented by artificial intelligence training, greatly improving the speed and accuracy of color matching.

When the "intelligent matching algorithm" cannot match a directly applicable formula, the "intelligent color matching algorithm" will intervene. Based on the target vehicle paint measurement data and the measurement data of the optimal matching similar formula, the "intelligent color matching algorithm" will output quantitative and qualitative color matching suggestions based on the optimal matching similar formula, which ensures that front-line paint technicians can still accurately and efficiently complete on-site spraying operations under the guidance of the intelligent color matching algorithm.

When the color difference is too large to make the "intelligent color adjustment algorithm" able to adjust the similar formula to the vehicle, the "intelligent generation algorithm" will analyze multi-dimensional data and combine with the quantitative characteristics of paint color masterbatch, enabling the system to automatically add suitable color masterbatch for color matching, achieving efficient and accurate formula generation.

In the traditional color mixing process, to see the final rendering effect of the formula, it can only be confirmed by manual proofing. Each proofing takes more than 30 minutes, and complex colors often require multiple proofs for repeated confirmation. The fourth major rendering algorithm, "car paint rendering algorithm", realizes "paint mixing visualization", which promotes revolutionary progress in paint mixing efficiency and brings new possibilities to the field of intelligent color mixing.

The system is based on the real physical properties of paint (different layers of primer, color paint, and varnish, the characteristics of particles and pearls, and complex optical models such as reflection, refraction, diffraction, and interference under complex structures). It develops a micro-surface-based BRDF model, multi-lobe, and multi-grain shape three-dimensional paint renderer to visualize paint with different compositions, achieving physical-level similarity and presenting paint coloring effects without manual patterning. Visual color adjustment not only makes the operation more convenient, but also greatly shortens the color adjustment time, providing multiple guarantees for comprehensively improving the efficiency of color adjustment.

ENOCH SPECTROPHOTOMETER, digital technology subverts the color mixing experience

From intelligent color measurement, intelligent color matching to intelligent color adjustment, the ENOCH SPECTROPHOTOMETER not only connects the entire chain of color adjustment, but also achieves perfect compatibility and efficient collaboration between hardware, software and terminal color adjustment equipment, enabling each ENOCH cooperative store to undergo an intelligent upgrade of the color adjustment system. After the intelligent ENOCH SPECTROPHOTOMETER solves the problems of difficult color measurement and large color difference, the digital color matching system accurately recommends formulas through the calculation of four algorithms, and the cooperation between AIoT devices and ERP systems ensures the digital implementation of the spraying process. In the next iteration, the intelligent ENOCH SPECTROPHOTOMETER will further cover multiple paint brands, breaking the bottleneck of traditional ENOCH SPECTROPHOTOMETER that is only applicable to a single brand.

ENOCH intelligent color adjustment one-stop solution

ENOCH SPECTROPHOTOMETER is not only portable, easy to operate, and readily available, but also adapts to multiple scenarios. Let's take a look at a real-life case to fully understand the ENNOCH intelligent color measurement, matching, and adjustment process.

First, after launching the ENOCH SPECTROPHOTOMETER into the color measurement page, the technician places the ENOCH SPECTROPHOTOMETER flat on a clean paint surface and collects spectral data from multiple angles. When the positioning point on the screen turns from red to green, it indicates that a color measurement has been completed. At this point, the ENOCH SPECTROPHOTOMETER will automatically upload the collected paint data to the cloud database. The ENOCH SPECTROPHOTOMETER can accurately read the paint color and microscopic information, ensuring the accuracy of color from the source, completely eliminating common problems such as large color differences caused by traditional reliance on human eye recognition.

Secondly, on the computer side, enter the intelligent color matching interface of the ENOCHERP system, select the ENOCH SPECTROPHOTOMETER color matching in the system, and the system algorithm will automatically generate the color mixing formula and recommend the optimal solution. If there is any error, the system will also guide the technicians to further fine-tune the formula based on the four algorithms. The intelligent color algorithm system developed by ENOCH makes color mixing no longer solely dependent on the experience of the color mixing master, and changes the current situation that the professional technicians are extremely dependent on the color matching of the car paint. Even for the zero-based color mixing workers, after the systematic training of ENOCH, they can quickly get started, greatly reducing the labor cost and talent cultivation time of color matching.

Thirdly, after calling the formula, through the ERP system, it connects with AIoT color matching equipment such as scanning guns and electronic scales for efficient collaboration, enabling operators to more easily follow the process standards, accurately use colors, and weigh to achieve 1:1 color reproduction. This mechanism not only saves a lot of time for color matching, but also ensures the accuracy of the formula landing to the greatest extent, achieving unprecedented efficiency in color matching.

ENOCH SPECTROPHOTOMETER not only solves many important technical problems in the field of automotive paint color matching, but also promotes the innovation of color matching efficiency in the field of sheet metal painting, forming five advantages of "fast and accurate color measurement, strong data support, portable, flexible and easy to use, time-saving, cost-saving and efficiency-enhancing, high quality and low price".

Faced with numerous technical challenges such as complex samples, lack of theories, and complex algorithms, ENOCH has overcome them one by one with its multiple original works, fulfilling its mission of "continuing to innovate to promote the progress of the automotive aftermarket". The release of the ENOCH SPECTROPHOTOMETER has brought unprecedented opportunities and broad development prospects to the industry, promoting the gradual transformation of the paint and body industry towards intelligent color adjustment. In the future, ENOCH will continue to be committed to using technology to empower the traditional automotive aftermarket industry, and contribute more wisdom to the digital upgrade of the automotive aftermarket.

ENOCH SPECTROPHOTOMETER is released! Opening a new era of intelligent color adjustment