Buying auto parts online is often a difficult and frustrating task. Auto parts are usually manufactured by third-party manufacturers, and quality and design are mixed. For example, if a 1995 Ford Fusion owner wants to order a new fender She may not be able to find the right product on the Internet. Even if a product is marked as 'compatible', due to manufacturing inconsistency, it is usually not suitable for this particular owner. Not only that, there are often hundreds of people on the market. The brand and model, which further increase the difficulty of the choice of the owner.
Amazon has already adopted AR technology in retail business and released specialized iOS and Android applications to help consumers visualize furniture and electronic products directly at home. According to a new patent discovered by EVNET, Amazon seems to want to expand AR. To the new retail category, AR technology applied to the automotive field.
Amazon began its active targeting of this vertical market in January 2017, when they announced that they had signed contracts with some of the largest auto parts suppliers in the United States. It is worth noting that due to conventional physical auto parts retailers such as AdvanceAutoParts and AutoZone on the official website or mobile The application does not provide support for AR functionality, so Amazon has found a breakthrough in this area.
In the patent entitled 'VehicleComponentInstallPreviewImageGeneration' (click to go), Amazon described the system to use AR to help users preview the image of auto parts.
Step 1: Use computer vision to identify vehicles
The system described in the patent will first identify the vehicle's brand, model and year through the user equipment camera (the application will store the vehicle's accurate measurement in the database).
The computer vision technology in the application program detects the vehicle based on the car battery, headlight or brand logo.
The computer vision technology in the application program detects the vehicle based on the car battery, headlight or brand logo.
If computer vision cannot detect the vehicle model, the user can choose to manually enter the vehicle's vehicle identification number.
After identifying the vehicle, the application program will determine the part of the vehicle that needs visualization and detect the access points of the detected vehicle connection points or selected car parts.
Step 2: Choose the Right Car Parts
After identifying the connection point of the vehicle, the user can select the desired car part (including accurate measurement) from the 3D rendering database, such as the supercharger.
The app can display boosters of different brands and sizes, allowing the user to switch between different options, as shown in the figure below.
Then, the user can preview and select the required car parts from the complete 3D rendered full online database.
Patent-explained application's internal recommendation engine, which is similar to the Amazon retail official website's recommended function.
Patent-explained application's internal recommendation engine, which is similar to the Amazon retail official website's recommended function.
For example, the recommendation engine can also provide supplementary parts that fit the same part of the car.
Step 3: Test in AR
After selecting the required car parts, the platform integrates the selected car part image with the vehicle image, as shown below. At this time, the user can determine whether the car part is suitable for the vehicle.
It is worth noting that the Amazon patent indicates that 3D renderings can determine whether a part fits into a vehicle within a certain fitness threshold, indicating that even with the most accurate measurements, there are still parts of the fitting that cannot support visual projection. (such as screws that connect parts to the vehicle).
In addition, the application also contains supplementary information such as consumer reviews and vendor reviews to provide some reference for your decision.
If AR technology can help Amazon accurately provide compatible parts, they will be able to help third-party auto parts industry gain more market share than the original distributors that provide more expensive (but guaranteed to be compatible) auto parts.
