Measuring Urban Sidewalk Practicability: A Sidewalk Robot Feasibility Index

Autonomous parcel delivery is attracting a lot of interest. Terrestrial delivery drones travel at lower speeds, are smaller and lighter than passenger cars. These features make them an ideal and valuable first step and experimental sandbox toward fully autonomous vehicles. To be useful, however, small wheeled drones need to operate on parts of the roads that are reserved to pedestrians. This is a challenge by itself. Pedestrian areas are less structured than road and abide by looser rules. The best route for a delivery drone may not be the shortest path; other aspects need to be accounted for that make a route more or less practical for the specific features of the vehicle. This paper introduces a quantitative analysis of these specific issues. The paper proposes a quantitative index that asses a route practicability for a small terrestrial drone. It combines different aspects that account for sidewalk width, sidewalk surface condition, route length and the number of driveways and crosswalks present on the way. We provide the mathematical definition of the index, and use our wheeled drone prototype to show how it can be used to classify and chose the best routes among a selection. Although the index is designed for autonomous drones, given the specific dynamic features of the drone, it can also be employed as is to quantify the accessibility of different routes for disabled people.