US 12,065,158 B2
Systems and methods for detecting an environment external to a personal mobile vehicle in a fleet management system
Ashley John Cooper, San Francisco, CA (US); Alison Marie Thurber, San Francisco, CA (US); Eahab Nagi El Naga, San Francisco, CA (US); Christy Fernandez Cull, Sunnyvale, CA (US); and Abdullah Ahsan Zaidi, San Diego, CA (US)
Assigned to Lyft, Inc., San Francisco, CA (US)
Filed by Lyft, Inc., San Francisco, CA (US)
Filed on Jul. 7, 2022, as Appl. No. 17/859,881.
Claims priority of provisional application 63/231,159, filed on Aug. 9, 2021.
Prior Publication US 2023/0092933 A1, Mar. 23, 2023
Int. Cl. B60W 50/14 (2020.01); B60W 40/06 (2012.01); G01S 13/86 (2006.01); G01S 13/89 (2006.01)
CPC B60W 50/14 (2013.01) [B60W 40/06 (2013.01); G01S 13/86 (2013.01); G01S 13/89 (2013.01); B60W 2050/146 (2013.01); B60W 2420/403 (2013.01); B60W 2420/408 (2024.01); B60W 2420/54 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A personal mobile vehicle (PMV) for riding on a road surface and managed by a fleet management system, the PMV comprising:
a frame carrying a wheel;
a radar sensor carried by the frame of the PMV, wherein the radar sensor is configured to (i) emit an electromagnetic wave signal towards an environment external to the PMV, and (ii) receive a radar reflection signal indicating a characteristic of the road surface of the environment external to the PMV; and
a processing unit communicatively coupled to the radar sensor, the processing unit being configured to:
determine one or more characteristic factors of the road surface of the environment external to the PMV according to the radar reflection signal, wherein the one or more characteristics factors include at least a signal mean power or signal attenuation of the radar reflection signal;
generate a road surface type classification distribution by using a first machine learning classifier to process the one or more characteristic factors of the road surface of the environment, wherein the road surface type classification distribution represents probabilities of the road surface belonging to one or more road surface types;
generate a vulnerable road user (VRU) classification probability distribution by using a second machine learning classifier, wherein the VRU classification probability distribution indicates whether a combination of the one or more characteristic factors is associated with the VRU;
determine, based on one or more of the road surface type classification distribution or the VRU classification probability distribution, that the road surface is associated with a predetermined road surface type of concern or the VRU;
in response to a determination that the road surface is associated with the predetermined road surface type of concern or the VRU, determine a recommended speed limit for the PMV to operate on the predetermined road surface type of concern; and
present an alert message indicating the recommended speed limit to a display associated with the PMV.