US 12,073,316 B2
Method for determining at least one indication of at least one change
Francesco Montrone, Riemerling (DE); Jan Wieghardt, Munich (DE); Marc Zeller, Munich (DE); and Bernhard Kempter, Munich (DE)
Assigned to SIEMENS AKTIENGESELLSCHAFT, Munich (DE)
Filed by Siemens Aktiengesellschaft, Munich (DE)
Filed on Oct. 29, 2019, as Appl. No. 16/667,053.
Claims priority of application No. 18203385 (EP), filed on Oct. 30, 2018.
Prior Publication US 2020/0134457 A1, Apr. 30, 2020
Int. Cl. G06F 8/65 (2018.01); G06F 30/27 (2020.01); G06N 3/08 (2023.01); G06N 20/00 (2019.01)
CPC G06N 3/08 (2013.01) [G06F 8/65 (2013.01); G06F 30/27 (2020.01); G06N 20/00 (2019.01)] 12 Claims
OG exemplary drawing
 
1. A method for determining at least one indication of at least one change to a hardware or software unit of an autonomous vehicle, the method comprising:
receiving at least one input data record having the at least one change and associated data;
determining the at least one indication of the at least one change by applying a learning-based approach to the at least one received input data record, the at least one indication indicating whether the at least one change impacts a safety-relevant or safety critical function of a vehicle control of the autonomous vehicle before the at least one change is implemented; and
implementing the at least one change to the hardware or software unit of the autonomous vehicle, in response to approving the at least one change;
wherein the learning-based approach is a trained neural network that is trained to distinguish whether the at least one change is relevant to the vehicle control of the autonomous vehicle using training data records that include a change, associated data, and an indication that indicates the change is relevant to safety.