US 11,995,565 B2
Road icing condition prediction for shaded road segments
Campbell D. Watson, Brooklyn, NY (US); Mukul Tewari, Lafayette, CO (US); Eli Michael Dow, Wappingers Falls, NY (US); and Levente Klein, Tuckahoe, NY (US)
Assigned to International Business Machines Corporation, Armonk, NY (US)
Filed by International Business Machines Corporation, Armonk, NY (US)
Filed on Apr. 7, 2020, as Appl. No. 16/842,134.
Prior Publication US 2021/0312306 A1, Oct. 7, 2021
Int. Cl. G06N 5/04 (2023.01); B60W 40/06 (2012.01); G01S 17/42 (2006.01); G01W 1/10 (2006.01); G06N 5/02 (2023.01); G06Q 50/26 (2024.01)
CPC G06N 5/04 (2013.01) [B60W 40/06 (2013.01); G01W 1/10 (2013.01); G06N 5/02 (2013.01); G06Q 50/265 (2013.01); G01S 17/42 (2013.01); G01W 2203/00 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for road condition prediction comprising:
selecting a road segment for road condition prediction based on weather conditions;
generating a solar radiation budget model for the road segment;
determining a vegetation model factor for the road segment;
updating the vegetation model factor, in a dynamic structures model for the road segment, based on growth prediction for the tree, wherein the growth prediction is based on biomass growth estimates for the tree and localized growth models for the tree based on long term climate conditions of the road segment;
updating the solar radiation budget model using a permanent structures model and the dynamic structures model, wherein the permanent structures model is based on static objects near the road segment and the dynamic structures model is based on vegetation near the road segment where the vegetation growth model is used to determine a location, height and size of the vegetation near the road to determine a level the road segment is shaded by the vegetation;
generating a road condition model for the road segment using the updated solar radiation budget model and weather variables to identify at least one of wet, snow covered, or ice covered road surfaces; and
outputting a road condition prediction for the road segment based on the road condition model.