Browse Prior Art Database

Cognitive Approach To Determine Problematic Vehicle Components Disclosure Number: IPCOM000250566D
Publication Date: 2017-Aug-03
Document File: 3 page(s) / 20K

Publishing Venue

The Prior Art Database

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 49% of the total text.

Cognitive Approach To Determine Problematic Vehicle Components


Two new Watson APIs are proposed which can be used by any automotive or robotic companies to detect the defective or malfunctioned components of any vehicle.


Cars will be taken to service station at a set time interval or after it has travelled a certain distance for a maintenance check, or sometimes due to a break down. Service center advisor will mark to carry out some common tasks like oil, air filter change,.etc. and also asks technical expert to thoroughly check the vehicle if there any abnormal or unusual issues that need to be addressed. Based on his knowledge or past experience he will guess/identify malfunction or wear tare of components and then suggest for either repair or replacement. Similar process takes places for industrial equipment and robotics. In some of the advanced machines, the components were equipped with sensors. These sensors collect the data and transfer the data to a remote system for analysis, based on the analysis the component will be declared as “needs repair” or “replacement” so as to avoid machine failures or vehicle breakdown. Having a sensor for each component is a costly affair, sometimes it may not be feasible to have sensor for a tiny components or bunch of tightly coupled components which work together. Most of the time experience of a technician plays a major role to judge the parts/components wear and tear or root causing an issue. Mechanical parts wear and tear or malfunction will be judged by experts(humans), either by considering the part or by carefully listening the noise which comes out of the component. This require strong human expertise. A new method is proposed to monitor and judge mechanical components or combination of components wear and tear or malfunction using cognitive engine. The current method uses cognitive engine like Watson, which will be trained with human expertise data of all the components and associated sounds/images. The trained data is used to classify the new component input and detect the defect associated with that component. We propose two new Watson APIs which can be used by any automotive or robotic companies to detect the defective or malfunctioned components of any vehicle. The advantages of this method are:

1. The problem with the component can be caught in the early stages when an unusual sound is generated from it. By catching a problem in early stage, we can fix the component quickly at less cost and reduce the impact of the damage.

2. By comparing the size, dimensions of each vehicle component over the time, the damage or wear/tear can be identified which can help the vehicle owner to take preventive action.

3. Known solutions user sensors to identify a problem but having sensor for each component is costly affair. The proposed idea will be cheaper alternative compared to sensor alternative.

This proposed method will talk about achieving human capability using neural networks...