IP Library › Granted Patent US 12,564,964
Granted Patent B2
US 12,564,964 · App. 17/451,169 · Granted Mar 3, 2026

Adaptive navigation using artificial intelligence for enhancing task performance in autonomous roaming robotic devices

Inventors: Jenny S. Li (Franklinton, NC); Nirmit V Desai (Yorktown Heights, NY); Dhiraj Joshi (Edison, NJ); Raghu Ramaswamy (Bangalore, IN); Satish Rajani (Dewas, IN)
Assignee: International Business Machines Corporation
B25J9/1697B25J13/006G05D1/0022G05D1/0221G05D1/0246
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,564,964
App. No.
17/451,169
Granted
Mar 3, 2026
Kind
B2
Abstract

Optimizing sensing capabilities of a roaming robotic device using Artificial Intelligence (AI) includes receiving data at a control system having a computer from a robotic device. The control system communicating a policy to the robotic device for choosing navigation actions for the robotic device. The received data is analyzed using the control system for determining when the received data meets a threshold for determining quality of the data. The analysis can include generating a model based on the received data where the model includes vector representation of inputs detected by a sensor array at the location. In response to the received data at the control system not meeting the threshold for determining quality, the robotic device communicating with the control system to collaborate in updating the policy to choose a next action.

Claims (47)

1 . A computer-implemented method for optimizing sensing capabilities of a roaming robotic device, comprising:

receiving data, at a control system having a computer, from a robotic device, the robotic device operating at a site using a policy received from the control system, the policy includes operating actions for the robotic device, the data being generated from sensors on the robotic device in response to a sensing task being performed by the robotic device;

analyzing the received data using the control system for determining when the received data meets a threshold for determining quality of the data, in response to the receiving of the data from the robotic device at the site;

the analysis including generating a model based on a sample set of data, the analysis also including vector representation of the received data as input data, and the analysis includes comparing the vector representation to the model;

determining when the received data does not meet the threshold for determining quality based on the analysis including the comparing of the vector representation of the input data to the model;

in response to the received data at the control system not meeting the threshold for determining quality, the robotic device communicating with an edge server, the edge server evaluating quality of the received data and determining that the quality of the received data does not meet the threshold, and the edge server and the robotic device collaborating in updating the policy to choose a next action, until the received data at the control system meets the threshold; and

communicating the next action from the edge server to the robotic device.

2 . The method of claim 1 , wherein the next action includes a navigation action including the robotic device moving to a next location in the site and initiating additional data collection by the robotic device for the sensing task.

3 . The method of claim 1 , wherein the next action includes the robotic device adjusting a camera on the robotic device.

4 . The method of claim 1 , wherein the next action is selected from a group consisting of: moving to a next location; adjusting a camera on the robotic device; adjusting a sensor on the robotic device; and waiting for a period of time.

5 . The method of claim 1 , further comprising:

creating an out of distribution detector (OOD) from the sample set of data to determine a distance between a distribution of data in the sample set of data, and the input data; and

using the distance in the determining when the received data does not meet the threshold for determining quality.

6 . The method of claim 1 , wherein the robotic device is a roaming edge device (RED) located at an edge of communications for roaming edge devices at the site.

7 . The method of claim 1 , wherein the threshold for determining quality includes a measured level of electronic noise which is acceptable for quality and thereby when the received data exceeds the threshold, the received data is unacceptable in relation to quality.

8 . The method of claim 1 , further comprising:

updating the model, using the computer; the updated model includes the following;

updating the received data;

updating the analyzing of the updated received data; and

communicating with the control system to collaborate in another updating of the policy for choosing a navigation action, in response to the updated received data not meeting the threshold for determining quality.

9 . The method of claim 8 , further comprising:

iteratively generating the model to produce updated models.

10 . The method of claim 1 , wherein the site where the robotic device is operating has low quality electronic wireless communication capabilities.

11 . The method of claim 1 , wherein the robotic device communicates with the control system at least in part as a cloud server to collaborate in updating the policy to choose a navigation action, and the cloud server being in communication with the robot at a location.

12 . A system using a computer for optimizing sensing capabilities of a roaming robotic device, which comprises:

a computer system comprising; a computer processor, a computer-readable storage medium, and program instructions stored on the computer-readable storage medium being executable by the processor, to cause the computer system to perform the following functions to;

receive data, at a control system having a computer, from a robotic device, the robotic device operating at a site using a policy received from the control system, the policy includes operating actions for the robotic device, the data being generated from sensors on the robotic device in response to a sensing task being performed by the robotic device;

analyze the received data using the control system for determining when the received data meets a threshold for determining quality of the data, in response to the receiving of the data from the robotic device at the site;

generate a model, as part of the analysis, based on a sample set of data, the analysis also including vector representation of the received data as input data, and the analysis includes comparing the vector representation to the model;

determine when the received data does not meet the threshold for determining quality based on the analysis including the comparing of the vector representation of the input data to the model;

in response to the received data at the control system not meeting the threshold for determining quality, the robotic device communicates with an edge server, the edge server evaluating quality of the received data and determining that the quality of the received data does not meet the threshold, and the edge server and the robotic device collaborate in updating the policy to choose a next action, until the received data at the control system meets the threshold; and

communicate the next action from the edge server to the robotic device.

13 . The system of claim 12 , wherein the next action includes a navigation action including the robotic device moving to a next location in the site and initiating additional data collection by the robotic device for the sensing task.

14 . The system of claim 12 , wherein the next action includes the robotic device adjusting a camera on the robotic device.

15 . The system of claim 12 , wherein the next action is selected from a group consisting of: moving to a next location; adjusting a camera on the robotic device; adjusting a sensor on the robotic device; and waiting for a period of time.

16 . The system of claim 12 , further comprising:

creating an out of distribution detector (OOD) from the sample set of data to determine a distance between a distribution of data in the sample set of data, and the input data; and

using the distance in the determining when the received data does not meet the threshold for determining quality.

17 . The system of claim 12 , wherein the robotic device is a roaming edge device (RED) located at an edge of communications for roaming edge devices at the site.

18 . The system of claim 12 , wherein the threshold for determining quality includes a measured level of electronic noise which is acceptable for quality and thereby when the received data exceeds the threshold, the received data is unacceptable in relation to quality.

19 . A computer program product for optimizing sensing capabilities of a roaming robotic device, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform functions, by the computer, comprising the functions to:

receive data, at a control system having a computer, from a robotic device, the robotic device operating at a site using a policy received from the control system, the policy includes operating actions for the robotic device, the data being generated from sensors on the robotic device in response to a sensing task being performed by the robotic device;

analyze the received data using the control system for determining when the received data meets a threshold for determining quality of the data, in response to the receiving of the data from the robotic device at the site;

generate a model, as part of the analysis, based on a sample set of data, the analysis also including vector representation of the received data as input data, and the analysis includes comparing the vector representation of the input data to the model;

determine when the received data does not meet the threshold for determining quality based on the analysis including the comparing of the vector representation to the model;

in response to the received data at the control system not meeting the threshold for determining quality, the robotic device communicates with an edge server, the edge server evaluating quality of the received data and determining that the quality of the received data does not meet the threshold, and the edge server and the robotic device collaborate in updating the policy to choose a next action, until the received data at the control system meets the threshold; and

communicating the next action from the edge server to the robotic device.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Oct 18, 2021
From: LI, JENNY S.; DESAI, NIRMIT V; JOSHI, DHIRAJ; RAMASWAMY, RAGHU; RAJANI, SATISH
To: INTERNATIONAL BUSINESS MACHINES CORPORATION
Reel/Frame 057815/0659 →
Continuity (1)
Related Publication 20230124038A1 · Apr 20, 2023
References Cited (38)
US 7663481B2 · Kim · 2010 [cited by applicant]
US 10136302B2 · Hu · 2018 [cited by applicant]
US 10455462B1 · Desai · 2019 [cited by applicant]
US 10618168B2 · Gros · 2020 [cited by examiner]
US 11082525B2 · Sethuraman · 2021 [cited by applicant]
US 20120170474A1 · Pekarske · 2012 [cited by applicant]
US 20150190925A1 · Hoffman · 2015 [cited by examiner]
US 20170048308A1 · Qaisar · 2017 [cited by applicant]
US 20180016780A1 · Cotter · 2018 [cited by applicant]
US 20190053148A1 · Lee · 2019 [cited by applicant]
US 20190188473A1 · Witt · 2019 [cited by examiner]
US 20190281132A1 · Sethuraman et al. · 2019 [cited by applicant]
US 20200073401A1 · Szatmary · 2020 [cited by examiner]
US 20200156242A1 · Imai · 2020 [cited by examiner]
US 20200177671A1 · Tofighbakhsh · 2020 [cited by applicant]
US 20200327371A1 · Sharma · 2020 [cited by applicant]
US 20210210217A1 · Xu · 2021 [cited by applicant]
US 20210232136A1 · Griffin · 2021 [cited by applicant]
US 20210243260A1 · Bhaumik · 2021 [cited by applicant]
US 20210256424A1 · Rajkumar · 2021 [cited by examiner]
US 20210278257A1 · Dharia · 2021 [cited by examiner]
EP 4419977A1 · 2024 [cited by applicant]
WO 2023066771A1 · 2023 [cited by applicant]
International Search Report and Written Opinion, International Application No. PCT/EP2022/078463, International Filing Date: Oct. 13, 2022, Date of Mailing: Feb. 9, 2023, 16 pages. [cited by applicant]
Minovski et al., “Modeling Quality of IoT Experience in Autonomous Vehicles”, IEEE Internet of Things Journal, vol. 7, No. 5, May 5, 2020, 18 pages. [cited by applicant]
Yuhas et al., “Embedded Out-of-Distribution Detection on an Autonomous Robot Platform”, Destion '21, May 18, 2021, arXiv:2106.15965v1, Jun. 30, 2021, 6 pages. [cited by applicant]
An, et al., “EiF: Toward Elastic loT Fog Framework for Al Services”, IEEE Communications Magazine, vol. 57 Issue: 5, May 2019, 7 pages, <https://ieeexplore.IEEE.org/document/8713796>. [cited by applicant]
Chen, et al., “Real-Time Object Tracking on a Drone With Multi-Inertial Sensing Data”, IEEE Transactions On Intelligent Transportation Systems, vol. 19, No. 1, Jan. 2018, pp. 131-139, <https://ieeexplore. eee.org/docume… [cited by applicant]
Christopoulos, et al., Adaptive Sensing for Instantaneous Gas Release Parameter Estimation, Proceedings of the 2005 IEEE International Conference on Robotics and Automation, 7 pages, <https://citeseerx.ist.psu.edu/viewd… [cited by applicant]
Deng, et al., “Edge Intelligence: The Confluence of Edge Computing and Artificial Intelligence”, arXiv:1909.00560v2, Feb. 10, 2020, 13 pages. [cited by applicant]
Knudson, et al., “Adaptive navigation for autonomous robots”, Robotics and Autonomous Systems, vol. 59, Issue 6, Jun. 2011, 32 pages. [cited by applicant]
Lee, et al., “NeuralFP: Out-of-distribution Detection using Fingerprints of Neural Networks”, 2020 25th International Conference on Pattern Recognition (ICPR), pp. 9561-9568, <https://ieeexplore.ieee.org/document/941248… [cited by applicant]
Liu, et al., “Client-Edge-Cloud Hierarchical Federated Learning”, arXiv:1905.06641v2, Oct. 31, 2019, 6 pages. [cited by applicant]
Loven, et al., “EDISON: An Edge-Native Method and Architecture for Distributed Interpolation”, Sensors 2021, 21, 2279, 20 pages, <https://doi.org/10.3390/s21072279>. [cited by applicant]
Mell, et al., “The NIST Definition of Cloud Computing”, National Institute of Standards and Technology, Special Publication 800-145, Sep. 2011, 7 pages. [cited by applicant]
Sanderson, et al., “Adaptive Visual Servo Control of Robots”, In: Pugh A. (eds) Robot Vision. International Trends in Manufacturing Technology, 1983, Abstract only, pp. 107-116, <https://link.springer.com/chapter/10.100… [cited by applicant]
Xu, et al., “Edge Intelligence: Architectures, Challenges, and Applications”, arXiv:2003.12172v2, Jun. 12, 2020, 53 pages. [cited by applicant]
Zhou, et al., “Working towards Adaptive Sensing for Terrain-aided Navigation”, 2019 International Conference on Robotics and Automation (ICRA), 8 pages, <https://www.researchgate.net/publication/335140925_Working_toward… [cited by applicant]