ACV Auctions has filed a patent for methods and systems to capture and process audio data of a vehicle engine. The invention involves a mobile device that records engine sounds and associates tags with the audio file to identify vehicle conditions. A server processes the audio file and provides access to it through an API. The system also includes a condition report server that retrieves the audio file and includes it in a vehicle condition report. Engine conditions can be detected and tagged using trained models. GlobalData’s report on ACV Auctions gives a 360-degree view of the company including its patenting strategy. Buy the report here.
According to GlobalData’s company profile on ACV Auctions, social media analytics was a key innovation area identified from patents. ACV Auctions's grant share as of September 2023 was 24%. Grant share is based on the ratio of number of grants to total number of patents.
The patent filed is for capturing and processing vehicle engine audio
A recently filed patent (Publication Number: US20230267780A1) describes a system for identifying engine conditions of a vehicle using audio recordings of the engine. The system includes a processor and a computer-readable storage medium with instructions for obtaining audio recordings of the engine during different operation segments, such as start, idling, load, and shut off. The system also collects metadata, including the vehicle identification number, and transmits the audio recordings and metadata to a computer for processing. The computer analyzes the audio recordings to identify engine conditions and associated confidence values, indicating the probability of detecting each condition in the recordings. The system then receives information indicating the engine conditions and confidence values via a communication network.
The patent also mentions additional features of the system. For example, the metadata can include information from the vehicle computer, providing further context for the analysis. The system can be integrated with a mobile device, which includes a microphone for capturing the audio recordings. The mobile device can also receive the information indicating the engine conditions. Additionally, the system may include a separate mobile computing device for receiving the information. To ensure accurate recordings, the system can be equipped with a case that dampens vibrations experienced by the mobile device.
The patent further describes a method for identifying engine conditions using machine learning models. The system receives audio recordings of the engine and a vehicle identification number via a communication network. The recordings are processed using machine learning models, such as deep convolutional neural networks, to obtain outputs indicating engine conditions and corresponding confidence values. The processing involves generating spectrograms or Mel spectrograms from the audio recordings and using them as inputs for the machine learning models.
In summary, the filed patent describes a system and method for identifying engine conditions of a vehicle using audio recordings. The system captures the recordings during different engine operation segments, collects metadata, and transmits the data for processing. Machine learning models are employed to analyze the recordings and provide outputs indicating engine conditions and confidence values. The system can be integrated with a mobile device and includes features to ensure accurate recordings.