Electric scooters (e-scooters) are proliferating rapidly as an inexpensive mode of transportation. GPS equipped smartphones are used to guide riders from points A to B, but GPS is known to have hundreds of meters of error in areas where the line of sight to navigation satellites is fully or partially obscured. To address this problem, we present ScootLoc, which enables precise e-scooter localization, by leveraging physical characteristics of sidewalk ramps, to correct for GPS error. We show that e-scooters equipped with a gyroscope and an accelerometer are able to uniquely identify sidewalk ramps and match the ramp to its physical location, thereby augmenting noisy GPS based navigation systems. We implement ScootLoc and present a preliminary evaluation on a route containing ten ramps, achieving 97% classification accuracy.