IBM’s Sherpa Drone: “Fly, my pretties, fly!”

There is a WW2 story about a boy who guided a munitions truck up a treacherous road to get to the front lines in the middle of a rainy night. The driver, unable to use headlights for fear of a sniper, followed the only thing visible to him: the boy’s white towel. This patent application is kind of like that, but instead of a boy it’s a drone, and the towel is the drone’s array of sensors.

Suppose there is an accident or a tree branch blocking a lane of travel. A UAV that is monitoring that part of the road sends a signal to the driver that there is a hazard up ahead, how far it is, and what to do about it. Ironically, “UAV program 110 may notify the user by sending a message to the user’s smart phone.”

Here’s the zinger: in the main version of the invention, the UAV is mounted on the user’s vehicle, and can be dispatched like a flying monkey in case of trouble. “UAV program 110 determines if a user’s view of the road is obstructed. If the user’s view is obstructed, UAV program 110 deploys UAV 108 to monitor the road for any impending hazards or danger.”

Query: won’t the driver realize that the view is obstructed, and what’s the threshold speed that makes the whole system inoperable? It’s unlikely that this is a shrouded military application, since the severity of hazards “can differ between a road that is completely blocked and a deer grazing by the side of the road.”

Here is a relevant paragraph: “UAV program 110 directs UAV 108 to fly ahead of the user’s vehicle to monitor for conditions that may be hazardous or dangerous. During the monitoring, UAV program 110 collects measurements about how far ahead a hazard is or for how long a hazard will persist. Hazards may include debris on the road, traffic jams, traffic accidents, the presence of police or emergency vehicles, and weather related road conditions, such as flooding, snow, fog, hydroplaning conditions, etc. For example, UAV program 110 may determine a traffic jam is up ahead by calculating the average distance between vehicles and the speed at which the vehicles travel and comparing the speed to the known speed limit for the road. UAV program 110 may direct UAV 108 to fly a certain distance ahead of the user’s vehicle. The distance may depend on the type of vehicle. For example, a small sedan can stop faster than a tractor trailer due to the weight of the vehicle, therefore the amount of warning time needed differs.”

Why IBM has gone in this direction is nebulous, and claim 1 is tailored to the description’s disclosure.

Claim 1. A method for hazard detection, the method comprising: determining, by one or more computer processors, whether an obstruction of view for a user in a first vehicle is detected; responsive to determining the obstruction is detected, deploying, by the one or more computer processors, a first unmanned aerial vehicle (UAV) associated with the first vehicle; and determining, by the one more computer processors, whether one or more hazards associated with a path of the first vehicle are detected.



US Patent Application No: 20160059962

Filed (US): September 3, 2014

Published: March 3, 2016

Applicant: International Business Machines Corporation


IBMsherpa IBMsherpa1