Vehicle sink alert system and method

Patent metadata

Field	Value
Publication number	US-11432452-B2
Application number	US-201916259411-A
Country	US
Kind code	B2
Filing date	Jan 28, 2019
Priority date	Jan 28, 2019
Publication date	Sep 6, 2022
Grant date	Sep 6, 2022

Abstract

A vehicle sink alert method and system that includes collecting image data, determining field conditions, calculating wheel sink depth, predicting vehicle sink events using wheel sink depth and field conditions, and activating an alert when a vehicle sink event is predicted. The method can include predicting vehicle sink events using GPS coordinates and historical vehicle sink data. The method can include determining and using vehicle motion to predict vehicle sink events. Predicting vehicle sink events can include forming a multi-dimensional model with positive sink points where vehicle sink events have occurred and negative sink points where vehicle sink events have not occurred; generating a current vehicle point using monitored and calculated data; determining whether the current vehicle point is within a positive sink region formed by the positive sink points; and predicting a vehicle sink event if the current vehicle point is within the positive sink region.

First claim

We claim:

1. A vehicle sink alert method for a vehicle in a surroundings, the vehicle having a plurality of wheels, the vehicle sink alert method comprising:
monitoring lateral force exerted on a monitored wheel of the plurality of wheels using a force sensor coupled to the monitored wheel, the force sensor providing lateral force readings;
collecting image data of the surroundings using a camera;
monitoring vehicle sensors providing vehicle parameter readings, the vehicle parameter readings including vehicle speed readings;
calculating a wheel sink depth for the monitored wheel using the lateral force readings and the vehicle speed readings;
calculating field conditions using the image data;
monitoring vehicle location;
retrieving historical vehicle sink data with associated vehicle sink locations for each historical vehicle sink event; and
predicting whether a vehicle sink event is expected using the calculated wheel sink depth, the calculated field conditions, the monitored vehicle location and the historical vehicle sink data; and
activating an alert device when a vehicle sink event is predicted;

wherein predicting whether a vehicle sink event is expected comprises:
forming a multi-dimensional data model comprising positive sink data points where a historical vehicle sink event has occurred and negative sink data points where a historical vehicle sink event has not occurred;
generating a current vehicle data point using the calculated wheel sink depth, the calculated field conditions, the monitored vehicle location and the historical vehicle sink data;
determining whether the current vehicle data point is inside a positive sink region formed by the positive sink data points or outside the positive sink region; and
predicting a vehicle sink event is expected if the current vehicle data point is inside the positive sink region.

2. The vehicle sink alert method of claim 1 :
wherein monitoring vehicle location comprises monitoring Global Positioning System (GPS) coordinates of the vehicle using a GPS sensor;
wherein the associated vehicle sink locations for each historical vehicle sink event comprises associated GPS coordinates for each historical vehicle sink event; and
wherein predicting whether a vehicle sink event is expected further uses the GPS coordinates.

3. The vehicle sink alert method of claim 1 , further comprising:
determining vehicle motion using the image data and the vehicle parameter readings, the vehicle parameter readings further including vehicle throttle position readings and vehicle gear position readings; and
wherein predicting whether a vehicle sink event is expected further uses the determined vehicle motion.

4. The vehicle sink alert method of claim 3 , further comprising:
calculating a wheel sink depth change for the monitored wheel using current and previous values of the calculated wheel sink depth;
wherein predicting whether a vehicle sink event is expected further uses the wheel sink depth change.

5. The vehicle sink alert method of claim 1 , wherein determining whether the current vehicle data point is inside or outside the positive sink region comprises:
estimating a boundary separating the positive sink data points and the negative sink data points;
determining that the current vehicle data point is inside the positive sink region when the current vehicle data point is on the same side of the boundary as the positive sink data points; and
determining that the current vehicle data point is outside the positive sink region when the current vehicle data point is not on the same side of the boundary as the positive sink data points.

6. The vehicle sink alert method of claim 1 , the vehicle parameter readings further including vehicle throttle position readings and vehicle gear position readings; and the method further comprising:
determining whether a current vehicle sink event is occurring at the current vehicle data point using the image data, the vehicle throttle position readings and the vehicle gear position readings.

7. The vehicle sink alert method of claim 6 , further comprising:
adding the current vehicle data point to the multi-dimensional data model as an additional positive sink data point if the current vehicle data point is inside the positive sink region or a current vehicle sink event is occurring at the current vehicle data point; and
adding the current vehicle data point to the multi-dimensional data model as an additional negative sink data point if the current vehicle data point is not inside the positive sink region and a current vehicle sink event is not occurring at the current vehicle data point.

8. The vehicle sink alert method of claim 1 , wherein calculating a wheel sink depth for the monitored wheel comprises:
detecting a submerge time when the lateral force readings for the monitored wheel increase above a typical range;
detecting an emerge time when the lateral force readings for the monitored wheel decrease back to the typical range;
calculating an elapsed time from the submerge time to the emerge time;
calculating a submerged circumference for the monitored wheel based on the elapsed time and the vehicle speed readings; and
calculating the wheel sink depth for the monitored wheel based on the submerged circumference for the monitored wheel.

9. A vehicle sink alert method for a vehicle in a surroundings, the vehicle having a plurality of wheels and a plurality of vehicle sensors, the vehicle sink alert method comprising:
receiving vehicle parameter readings from the plurality of vehicle sensors;
calculating a wheel sink depth for a monitored wheel of the plurality of wheels using the vehicle parameter readings;
receiving image data of the surroundings from a camera;
calculating field conditions using the image data;
predicting whether a vehicle sink event is expected using the calculated wheel sink depth and the calculated field conditions; and
activating an alert when a vehicle sink event is predicted;

wherein predicting whether a vehicle sink event is expected comprises:
forming a multi-dimensional data model comprising positive sink data points where a historical vehicle sink event has occurred and negative sink data points where a historical vehicle sink event has not occurred;
generating a current vehicle data point using current vehicle parameters;
determining whether the current vehicle data point is inside a positive sink region formed by the positive sink data points or outside the positive sink region; and
predicting a vehicle sink event is expected if the current vehicle data point is inside the positive sink region.

10. The vehicle sink alert method of claim 9 , further comprising:
receiving Global Positioning System (GPS) coordinates of the vehicle;
retrieving historical vehicle sink data with associated GPS coordinates for each historical vehicle sink event; and
wherein predicting whether a vehicle sink event is expected further uses the GPS coordinates and the historical vehicle sink data.

11. The vehicle sink alert method of claim 10 , further comprising:
calculating a wheel sink depth change for the monitored wheel using current and previous values of the calculated wheel sink depth;
wherein predicting whether a vehicle sink event is expected further uses the wheel sink depth change.

12. The vehicle sink alert method of claim 9 , further comprising:
determining vehicle motion using the image data and the vehicle parameter readings, the vehicle parameter readings including vehicle throttle position readings and vehicle gear position readings; and
determining whether a current vehicle sink event is occurring at

Assignees

• Deere & Co

Inventors

• Balani Kirti
• Krishnakumar Shibinkumar
• Rane Vishal

Classifications

• A01B69/004

  automatic · CPC title

• G06N20/00

  Machine learning · CPC title

• B60C23/0477

  Evaluating waveform of pressure readings · CPC title

• G06Q50/02 Primary

  Agriculture; Fishing; Forestry; Mining · CPC title

• B60C23/0408

  transmitting the signals by non-mechanical means from the wheel or tyre to a vehicle body mounted receiver · CPC title