Skip to content

Possible Scope of the Project.

Jump to bottom
prithvisekhar edited this page Mar 27, 2020 · 8 revisions

The list of features can be viewed here

In the existing system of automotive, we have MIL(Malfunction Indicator Lamp) which indicates a malfunction in the system, It has two stages: steady (indicating a minor fault) and flashing (indicating a severe fault), So the drawback into this system is that MIL gives an indication only after the malfunction has occurred, another drawback is the lack of remote access to the vehicle data.

To deal with these issues, a system can be developed to predict the health of the vehicle even before the fault occurs. To achieve this there are 4 stages:

Capture Vehicle data. Send captured data to a central device through sensor networks. Push data unto a Cloud Platform. Perform Prediction Analysis using ML techniques.

Possible Scope of the project.

1. Police Visions

  1. Detecting Traffic Signal and its violation.
  • Traffic Signal Light Detection.
  • Extracting Number Plates of Vehicles.
  • Detection of the violation act and make an informed decision.
  1. Moving in Opposite Direction in Lane and false parking.
  • Detecting Number Plates of Vehicles.
  • Detecting and putting down the path and movement of the vehicle in a programmable form.
  • Making decision and storing evidence of the same with respect to the tickets to be generated.
  1. Helmet Detection and Seat Built [Depending on the picture clarity]. i. Detection of 2-wheeler
  • Detecting the upper portion of the rider and detection of helmet.
  • Making decision based on the vehicle movement and present helmet. ii. Detection of 4Wheeler
  • Only if the image is clear. We can detect the if seat belt is present or not
  • Accordingly, decision can be taken based on the vehicle motion and their seatbelt position.
  1. Speed
  • Detecting of the speed of the vehicle.
  • Depending on the location finding the speed limit at that GPS location and decide accordingly.
  1. Traffic Sign Violation
  • Detection of the Traffic Sign.
  • Check the response of the vehicle based on the sign.
  • Decide on the violation of the same. And log it for evidence of the court.
  1. Police notification for road held and ambulance path overlap. i. As per the requirement. The police can give road withhold (Due to VIP etc.) data on the given platform. ii. Find the data routing path for ambulance for clearance of shortest path. Thus, also adjusting the traffic lights accordingly as it passes.
  • Route Detection and mapping the shortest route from source to destination.
  • Sending the signal to all the platform and synchronizing the traffic light signal. iii. Construction sites would also have similar adjustments.

2. Government update on Traffic, Road Profile and Possibility of changes in the future road construction planes.

  1. The road profile used for fixing the damage. i. The degree and importance of the given situation based on

    • Usability of the road.
    • Size and degree of damage. ii. Higher damages like flyover damage.
    1. Degree of damage based on accident and flyover images.
    • Create data base over the past data sets.
    • And have a decision algorithm, which can predict the cause as flyover damage.
    • Have flyover image to verity the conditions of the flyover.
    1. Traffic Impact and altering of the situation for the same.

  2. Traffic Light Timings i. Usually the traffic light time are static in nature. But according to the situation and the history and other parameters it can be made dynamic in nature.

    1. Coming up with predictive algorithm. a. Deciding the feature and parameters for the output of the model. b. Creating a predictive model for the same. Can also have an UBM Model for the same. c. Run our data set to find the likelihood in that model.
    2. Interrupts in the algorithm as per the situation. a.Ambulance and VIP b.Construction in an Area c.Other Interrupts etc.

  3. Long Term Construction Plans of the Government. i. The planning of government of various activities like 1.Contruction of flyovers 2.Widening of road 3.Creation of new transport station. etc. All these planes can be decisively made as per the history of the general road and traffic profile and other parameters. a. Decide long term effecting traffic parameters. b. Averaging and cumulating that parameters to creating a predictive model for the same.

3. Time and Navigation path for traffic light control and time synchronization.

  1. Time Management of daily travelers for a given Source and Destination to reduce traffic flow. i. Getting data of all the Companies, Schools and other daily traveler's data along with path details. So that the respective organization can come together to have a distinct timing and path. Many solutions can be formulated to reduce the traffic flow. The following parameters can be considered 1.Time and Shift Difference. 2.Different Route Allocation. 3.Vehicle Pooling. ii. Traffic Light Control 1. Based on the prerequisite knowledge of the travelers on the road the Traffic Light can also have these parameters to dynamically change its switching times. iii. Suggestion of mode of transport.
    1. Cost Based a. The algorithm should be able to give least possible cost for a given source destination pair. b. And accordingly predict the time.
    2. Time Based a. Least time taking route and means of transport (including bus,train,metro,car,bike, etc.) b. Providing a detailed cost and time analysis for the same. 3. Time and Cost a. Merging the both parameter and having a threshold of both to consider that parameter at that much weightage.

4. Live traffic congestions

  1. Live Feeds of data and interpretation of the current traffic scenarios and the prediction of the traffic in future timing (Time change of about 1-6 hr. difference also should be shown) i. Time Change would be like a parameter change keeping 15 [Changeable according to feasibility] as step sizes. ii. All Ambulance and VIP movement should be updated as fast as possible as per the information reaches the authorities. iii. Construction and road profile can also be a live feed. To changes paths accordingly.
  2. Live Feed can also have an overhead algorithm for ambulance. To predict which lane can be most effectively cleared for the ambulance route to go by.

5. Crowdsourcing

  1. Some ethical and legal obligation might hinder in this application. But we can do the following without considering it as of now. i. Delivery of things can be shifted as travelers picking and dropping things. Which can be synchronized and scheduled by the algorithm. [Cut cost and traffic time as well] 1. Time Routing algorithms 2. Shortest Route Algorithm 3. Decision making for the routed path. ii. Postal, Courier and Food Delivery Service can be synchronized for the same to minimize and eliminate the traveler involvement in the process.

6. Vehicle Diagnostic

  1. Estimating remaining useful life of a system
  2. Vehicle health monitoring (Fleet management)
  3. Predictive maintenance
  4. Fuel trims
  5. Remote monitoring of vehicle data (Can be used by insurance companies)
  6. Predicting Car price depending upon its condition
  7. Emission monitoring
  8. Headlight intensity
  9. Alerting nearest mechanic in case of breakdown
  10. Summary of car before starting the journey
Clone this wiki locally