Scooting around geolocation problems

Here at Nestwave, we are excited about the future of geolocation and our next generation GNSS and hybrid GNSS + 4G + WiFi solutions. We’re working to dramatically move the needle for IoT devices in the areas of power consumption, accuracy and indoor coverage.

Many of us are familiar with the reliability problems of geolocation when using our smartphone. When walking or driving in an urban area, it remains common that our phones think we’re a half a block or more up the street, sometimes even locating us in the middle of a building. Once we step indoors or enter an underground parking structure, we often lose location entirely. Turns out, the problems (and implications) are critically magnified when we factor in the additional constraints and requirements associated with new IoT applications.

Let’s take the example of dockless scooters, a micro-mobility solution familiar to many of us residing in urban centers. Paris alone has 20,000 scooters in operation. Scooters from Lime, Bird and others are equipped with geolocation solutions based on GNSS satellites with some using additional signal sources such as WiFi to build an ensemble map of the most likely scooter position. Problem solved, right? Not so fast.

The most visible use of scooter location is the pin map showing customers the nearest available scooters using a service provider’s app. This allows each of us to quickly find and rent a scooter, in real time. However, scooters operate and are often left in a variety of environments, and often in places particularly challenging to obtain an accurate geolocation fix – urban areas with tall buildings, parking garages, bridge underpasses and even in building lobbies. The effect is that customers waste time looking for scooters while companies lose revenue opportunities and damage to their reputation. Examples of this frustration abound on the internet, including here and here.

Rather less obvious are two location-based issues that are even more critical to the long term success of micromobility.

First, is the location problem involved in the collection, charging and release of scooters. Scooter companies such as Bird and Lime pay freelancers to find, recharge and relocate scooters on a nightly basis.

The backbone of the solution is the location reported by the scooter used by freelancers to find each scooter. Due to poor GNSS performance, this location is often incorrect, leading workers to spend time walking and driving to find the missing scooters. The problem is so pervasive, there are entire forum posts and blogs (here, here, here) dedicated to the problem.

There is an environmental impact as well. With 40% of a scooter’s CO2 footprint coming from the pickup/return service, the effect of inaccurate location is substantial. The 20,000 scooters in Paris result in over 4.7 million km being driven each year to pick up, recharge and return scooters to service. Improving the location accuracy on just 5% of these trips would result in a CO2 reduction of almost 30,000 kg per year.

Perhaps the most acute problems that location accuracy has on the dockless scooter business relate to safety and policy. City governments around the world face intense pressure to manage and regulate micromobility to address the growing problems and risks around use, including safety, liability and property rights. Injuries from scooter use – both drivers and pedestrians – is a growing problem.

Cities are attempting to control scooter use by creating speed restricted and exclusion zones, an approach called geofencing. But here again, poor location accuracy is causing significant problems. For one, the accuracy is not good enough to distinguish between areas that are nearby to each other – a street vs a sidewalk, or a park vs a bike path. This leads to unpredictable, automatic changes in scooter behavior (e.g. speed limiting) and has resulted in customers simply ending the ride and walking away in frustration.

Conversely, location errors of just 5-10m can allow a scooter to enter a restricted zone at speed, frustrating and endangering groups of people in locations where scooters just don’t make sense (densely populated public plazas, pedestrian-only walkways).

According to a CNN article “organizations who have requested restrictions described the GPS technique as ineffective”

Speed is determined on a scooter based on location samples. Depending on the battery level, the update rate may be very much reduced, dramatically affecting speed accuracy and thus improperly triggering changes to the scooter behavior (e.g. speed limiting). There are even examples cited where this can cause a safety issue – while crossing a street or going around a car.

These are just a few examples of how critical geolocation performance is to emerging IoT applications. For micromobility, location accuracy, indoor/underground capability and low power performance plays a key, central role in the usability, safety, policy and economics of solutions.

At Nestwave, we will continue to work on new and better ways to bring improved location accuracy (3-5x), reduced power dissipation (1/10th) and indoor capabilities to the next generation of IoT devices.

Perhaps we’ll be seeing you on an electric scooter soon!

Other posts from our blog…