Granted Patent B2

US 12,517,260 · App. 17/894,660 · Granted Jan 6, 2026

Inventors: Jonathan Fetter-Degges (Washington, DC); John Karvounis (Bowie, MD); Carole Teolis (Glendale, MD)

Assignee: TRX SYSTEMS, INC.

G01S19/21G01S19/08G01S19/37G01S19/396G01S19/45

Abstract

Claims (27)

1 . A method for automatically adjudicating sensor input for use in a navigation solution, comprising:

identifying N potentially compromised sensor inputs, generating a first navigation solution including trusted sensor input and without the N potentially compromised sensor inputs, and outputting the first navigation solution; and

utilizing one or more levels of integrity monitoring, including computing a state divergence value within a state space filter based on changes in uncertainty after sensor input is applied, to determine if one or more of the N potentially compromised sensor inputs can be validated, generating at least a second navigation solution including the trusted sensor input and with one or more of the N potentially compromised sensor inputs, and

based on a validation, automatically eliminating non validated sensor inputs and outputting the second navigation solution as a position and map estimate for use in a navigation system, including applying the second navigation solution to constrain further state updates within the state space filter.

2 . The method of claim 1 , wherein the first navigation solution and the second navigation solution are generated by a particle filter that contains a first set of particles that do not apply the N potentially compromised sensor inputs and a second set of particles that apply one or more of the N potentially compromised sensor inputs.

3 . The method of claim 2 , wherein the utilizing one or more levels of integrity monitoring includes maintaining first integrity values for sensor inputs in the first set of particles and second integrity values for the second set of particles and correlating the first integrity values and the second integrity values from new sensor input to new sensor input throughout a duration of a period.

4 . The method of claim 1 , wherein the first navigation solution and the second navigation solution are generated by a Kalman filter, wherein the Kalman filter does not apply the N potentially compromised sensor inputs to derive the first navigation solution, wherein the N potentially compromised sensor inputs are applied to one or more additional navigations solutions including the second navigation solution, wherein each of the one or more additional navigation solutions applies a different weight based on application of the N potentially compromised sensor inputs.

5 . The method of claim 1 , wherein the utilizing one or more levels of integrity monitoring includes a plurality of navigation solutions, wherein each navigation solution among the plurality of navigation solutions includes the trusted sensor input and a different sensor input among the N potentially compromised sensor inputs and includes a full navigation solution that includes the trusted sensor input and each of the N potentially compromised sensor inputs.

6 . The method of claim 5 , wherein the full navigation solution is utilized as the navigation solution when all of the sensor inputs are validated.

7 . The method of claim 5 , wherein the full navigation solution is eliminated and rewound including only the trusted sensor input when all of the sensor inputs are not validated.

8 . A computer-implemented method for automatically adjudicating sensor input for use in a navigation solution, comprising:

receiving, by a computing device comprising a processor and memory, input data from a sensor and a preprocessed integrity estimate of the input data;

executing, by the computing device, a state space filter configured to generate a navigation solution based on the input data;

splitting the navigation solution within the state space filter into a trusting solution and an untrusting solution based on the preprocessed integrity estimate;

performing, by the computing device, a cross consistency check of the trusting solution and the untrusting solution based on input data from at least one additional sensor;

adjusting, by the computing device, the state space filter based on the cross consistency check;

computing, by the computing device, a divergence within the state space filter based on changes in uncertainty after sensor input is applied, wherein the divergence comprises at least one of a difference in estimated position, velocity, or uncertainty covariance;

automatically determining, by the computing device, whether the divergence exceeds a predetermined threshold, and in response, updating a trust score associated with the sensor input;

adjusting, by the computing device, a weighting of the sensor input in the state space filter based on the updated trust score;

fusing all, by the computing device, part or none of the trusting solution and the untrusting solution into the navigation solution based on the updated trust score;

based on the updated trust score indicating the sensor input is not validated, automatically eliminating the sensor input from participation in the navigation solution; and

outputting, by the computing device, the navigation solution as a position and map estimate for use in a navigation system, including applying the navigation solution in subsequent navigation processing by adjusting the state space filter and constraining subsequent state updates based on the navigation solution.

9 . The method of claim 8 , wherein the state space filter is a particle filter that contains a first set of particles corresponding to the untrusting solution and a second set of particles corresponding to the trusting solution.

10 . The method of claim 9 , wherein the performing a cross consistency check includes maintaining first integrity values for sensor inputs in the first set of particles and second integrity values for the second set of particles and correlating the first integrity values and the second integrity values from new sensor input to new sensor input throughout a duration of a period.

11 . The method of claim 9 , wherein the adjusting the state space filter based on the cross consistency check and the adjusting the state space filter based on the divergence includes adjusting a first particle distribution of the first set of particles and a second particle distribution of the second set of particles.

12 . The method of claim 8 , wherein the state space filter is a Kalman filter, wherein the Kalman filter does not apply potentially compromised sensor inputs to derive the untrusting solution, and applies the potentially compromised sensor inputs to derive the trusting solution.

13 . The method of claim 12 , wherein the adjusting the state space filter based on the cross consistency check and the adjusting the state space filter based on the divergence includes adjusting a solution weight of the untrussing solution and the trusting solution.

Assignments (3)

Continuity (2)

References Cited (129)