Granted Patent B2

US 12,524,706 · App. 18/379,770 · Granted Jan 13, 2026

Inventors: Om Purushotham Akarapu (Hyderabad, IN); Daniel D. Widjanarko (San Francisco, CA); Durgadatta Belhekar (East Mumbai, IN); Jitendra Upadhyay (Bhopal, IN); Rama Krishnam Raju Rudraraju (Hyderabad, IN)

Assignee: Bank of America Corporation

G06N20/00G06F16/90335G06N3/08

Abstract

Claims (52)

1 . A sensor computing system, comprising:

one or more processors; and

a memory storing computer-readable instructions that, when executed by the processor, cause the system to:

receive sensor data from one or more sensors;

preprocess, by a simplified state estimation (SSE) module comprising a graphics processing unit (GPU) of the sensor computing system, the sensor data to calculate a variance, wherein latency of the received sensor data is reduced;

generate a processing code for the sensor data based on the variance, wherein the processing code changes based on a change in the variance;

associate the processing code with the sensor data;

associate a prioritization code with the sensor data, wherein the processing code and prioritization code have a one-to-one relationship where no two prioritization codes are used for the same processing code and no two processing codes are used for the same prioritization code;

translate the sensor data by an Internet of Things gateway device into a data format for processing by a real-time processing platform, wherein the Internet of Things gateway device communicatively couples the sensor computing system to the real-time processing platform; and

send the sensor data, the processing code, and the prioritization code to the real-time processing platform, wherein the real-time processing platform processes the sensor data, when the platform is busy, in an order based on at least the prioritization code and wherein the platform reads the processing code and executes a program corresponding to the processing code.

2 . The system of claim 1 , further comprising the one or more sensors, wherein the one or more sensors are communicatively coupled to the processor and comprise one or more of:

an ultrasonic sensor;

a camera;

a microphone;

a pressure sensor; and

a passive infrared receiver.

3 . The system of claim 1 , further comprising:

a neural network comprising a graphics processing unit configured to calculate the variance using the sensor data and an output of the SSE module.

4 . The system of claim 1 , wherein the SSE module uses a machine learning model trained on a dataset to calculate the variance.

5 . The system of claim 1 , wherein the processing code corresponds to one of: data fusion, remove noise, replace missing data, and historical data access.

6 . The system of claim 5 , wherein the one or more sensors are arranged in an automated teller machine, and the sensor computing system designates the received sensor data for a processing code corresponding to historical data access.

7 . The system of claim 5 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to historical data access, query a database for historical sensor data sharing characteristics with the received sensor data.

8 . The system of claim 5 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to remove noise, reduce or enhance the sensor data by removing noise.

9 . The system of claim 5 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to data fusion, receive additional sensor data from the one or more sensors.

10 . The system of claim 5 , wherein the one or more sensors are arranged in a parking lot of a retail banking location, and the sensor computing system designates the received sensor data for data fusion.

11 . The system of claim 1 , wherein associate the processing code with the sensor data comprises inserting the processing code into a first field in a header of a packet storing the received sensor data, and

wherein associate the prioritization code with the sensor data comprises inserting the prioritization code into a second field in the header of the packet.

12 . A method comprising:

receive, by a sensor computing system comprising at least a processor and a memory, sensor data from one or more sensors;

preprocess, by a machine learning model trained on a dataset, the sensor data to calculate a variance;

generate, by the processor of the sensor computing system, a processing code for the sensor data based on the variance, wherein the processing code changes based on a change in the variance;

associate, by the processor of the sensor computing system, the processing code with the sensor data;

associate, by the processor of the sensor computing system, a prioritization code with the sensor data, wherein the processing code and prioritization code have a one-to-one relationship where no two prioritization codes are used for the same processing code and no two processing codes are used for the same prioritization code;

translate the sensor data by an Internet of Things gateway device into a data format for processing by a real-time processing platform, wherein the Internet of Things gateway device communicatively couples the sensor computing system to the real-time processing platform; and

send the sensor data, the processing code, and the prioritization code to a real-time processing platform, wherein the real-time processing platform processes the sensor data, when the platform is busy, in an order based on at least the prioritization code and wherein the platform reads the processing code and executes a program corresponding to the processing code.

13 . The method of claim 12 , wherein the one or more sensors comprises one or more of: an ultrasonic sensor, a camera, a microphone, a pressure sensor, and a passive infrared receiver.

14 . The method of claim 12 , wherein associate the processing code with the sensor data comprises inserting the processing code into a first field in a header of a packet storing the received sensor data; and

wherein associate the prioritization code with the sensor data comprises inserting the prioritization code into a second field in the header of the packet.

15 . The method of claim 12 , wherein the processing code corresponds to one of: data fusion, remove noise, replace missing data, and historical data access.

16 . The method of claim 15 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to historical data access, query a database for historical sensor data sharing characteristics with the received sensor data.

17 . The method of claim 15 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to remove noise, reduce or enhance the sensor data by removing noise.

18 . The method of claim 15 , wherein the real-time processing platform is configured to, in response to the processing code corresponding to data fusion, receive additional sensor data from the one or more sensors.

19 . One or more non-transitory computer-readable media storing instructions that, when executed by a sensor computing system comprising at least one processor and memory, cause the sensor computing system to:

receive sensor data from one or more sensors;

preprocess, by a simplified state estimation (SSE) module comprising a graphics processing unit (GPU), the sensor data to calculate a variance;

generate a processing code for the sensor data based on the variance, wherein the processing code changes based on a change in the variance;

associate the processing code with the sensor data;

associate a prioritization code with the sensor data; wherein the processing code and prioritization code have a one-to-one relationship where no two prioritization codes are used for the same processing code and no two processing codes are used for the same prioritization code;

translate the sensor data by an Internet of Things gateway device into a data format for processing by a real-time processing platform, wherein the Internet of Things gateway device communicatively couples the sensor computing system to the real-time processing platform; and

send the sensor data, the processing code, and the prioritization code to a real-time processing platform, wherein the real-time processing platform processes the sensor data, when the platform is busy, in an order based on at least the prioritization code and wherein the platform reads the processing code and executes a program corresponding to the processing code, and

wherein latency of the received sensor data is reduced after execution of the instructions stored in the non-transitory computer-readable media of the sensor computing system.

20 . The one or more non-transitory computer-readable media of claim 19 , wherein the processing code corresponds to one of: data fusion, remove noise, replace missing data, and historical data access.

Assignments (1)

Continuity (2)

References Cited (29)