Another driverless vehicle on the streets: what should we learn from this?
The driver of an Uber self-driving car, who caused a pedestrian fatal accident in May 2018, was charged with negligent murder last month. Although the main reason for this accident was that the backup driver of the car was watching a TV show at the time of the accident, drivers of autonomous vehicles are often found sleeping in moving cars or doing other things with both hands. Of course, there is a social consensus that self-driving cars need to be regulated, but human drivers and pedestrians are still exposed to the tragic mistakes of inexperienced self-driving cars. Nevertheless, the fact that driverless e-scooters launched urges another discussion of another driverless vehicle.
Similar to what Uber started self-driving on real roads with Volvo, the self-driving e-scooter, which combines GoX’s e-scooter with Tortoise’s technology, has launched commercial service in Georgia, USA. Tortoise, a startup founded by a former Uber executive, does not launch autonomous e-scooters but rather seeks to become an operating system for shared micro-mobility vehicles and to make all types of shared vehicles autonomous. Equipped with $100 worth of gears, including two cameras, extra wheels and a radar, any shared e-scooter can drive itself. Tortoise’s co-founder aims to facilitate building the next generation of transport for all riders and cities with this “Groundbreaking” technology. However, there are still challenges to be faced.
Uncertainty of remotely controlled vehicles
To prevent mistakes in early self-driving cars that are not smart enough, it is not that Tortoise drives completely autonomously. The 100 commercially available e-scooters are actually remotely controlled by employees located in Mexico City and are only partially autonomous, highlighting that they are not 100% self-driving scooters. But, it is questionable whether the remote controlling can guarantee that the scooter on-site and its surroundings are safe. According to the results of an analysis of military drone collision data for 10 years from 2009, the main causes of remotely controlled unmanned drone accidents were power off due to electrical failure, wrong decisions made by pilots, full autonomy by lost link, poor weather, and a bird strike. There is no reason why most of the mentioned causes of drone accidents are not possible dangers while e-scooters move remotely without a driver. Besides, e-scooters can face more obstacles, such as dogs and cats, and can get stuck in narrow streets with a high probability of disconnection. If a micro-mobility vehicle intended to travel the ‘last mile’ that public transport does not cover, the possibility of disconnection is even higher while moving outside the city centre and into the suburbs. Then, is this another new social experiment for remotely controlled vehicles?
Uncertainty of low-speed autonomy
Tortoise moves at a low speed as its name so that the driverless scooter can stop not too late when encountering unexpected obstacles. Its co-founder’s claim that low-speed autonomy should take priority over high-speed autonomy seems to make sense, but this doesn’t mean that slow-moving autonomous vehicles should be on the streets as they are safe. Also, it is unclear who the safety of low-speed autonomy is for, obscuring the path that is suitable for e-scooters. Cities that have already introduced e-scooters tend to let them use the roads like bicycles which have the most similar speed. But won’t the driverless e-scooters interfere with other scooters or bikes with drivers when driving at low speeds? Because of this risk, Tortoise enables self-driving modes on mainly empty sidewalks and roads. This means that e-scooters equipped with Tortoise gear can become pedestrians when there is no driver, or they can be bicycles or vehicles as well. Could this be an additional advantage Tortoise offers exclusively to an e-scooter that no one controls? Low-speed autonomy can be compared to high-speed autonomy, but it is not clear who will be able to allow them to use sidewalks.
Uncertainty of optimised routes
Dockless e-scooter falling down and being parked on the road is another problem. Tortoise uses two aid wheels to prevent the self-driving e-scooter from falling over on rough roads, and optimises by monitoring smooth paths that are not crowded as much as possible. A driverless shuttle on the university’s campus shows that autonomous is suitable for relatively short or simple routes, and repeatedly travels on a fixed route. Hence, driving on optimised paths can contradict the characteristics of an e-scooter and its purpose of autonomous. One of the purposes of e-scooter autonomous driving for users is to automatically send people the e-scooter where they need it. Naturally, the demand for e-scooters will increase in crowded areas, and e-scooters that have finished travelling outside the city centre are likely to be called back to crowded areas. Additionally, when a user calls a taxi with an app, it’s easy to understand the roads that cars cannot enter, but it’s difficult to think of roads where scooters cannot enter. That is, considering the characteristics of micro-mobility, driving an optimised route avoiding crowded and complex places will limit the user’s convenience.
Who really needs an autonomous e-scooter?
The answer to this question can be found in the shared e-scooter market. Shared e-scooters, similar to bicycles, has been introduced as an eco-friendly alternative to traditional cars. But, every night there was a need for companies to collect, charge, and redistribute scooters scattered throughout the city by vans, and many devices were damaged in the process of fierce competition of scooter hunters. Consequently, the life of e-scooters was shortened, requiring shared scooter providers to spend high operating costs. Moreover, considering the manufacture, disposal and maintenance of e-scooters, including fossil fuel-based vans, it was estimated that they would emit more greenhouse gases than buses or electric cars. The reason shared scooter service providers in Paris have contracted with scooter collectors using electric vans, invested in the development of replaceable batteries and improved their business environment is to reduce their environmental impact. A platform automatically repositioning e-scooters reduces the need for e-scooter companies to pay for scooter collection companies. Tortoise, which requires relatively low fees, may be an innovative way to significantly reduce the cost of collecting scooters. Will the amount of money that users of self-driving shared scooters have to pay also dramatically decrease? We’ll have to see this. Otherwise, this would be just an opportunity for shared e-scooter companies to increase their profits.
How about the UK?
Fortunately, people can’t see the driverless e-scooters on the streets in the UK yet, but e-scooter rentals started being allowed experimentally to avoid crowded public transport when easing the lockdown. Shared e-scooters licensed by their cities can now use roads and dedicated cycle routes similar to bicycles, during the trial. Right, although e-scooters are already widely sold across the UK and are more hygienic than sharing them, private e-scooters were not included in this trial. This may be a part of government strategies to minimise risks and maximise benefits before legalising this new mode of transport on a large scale. Efforts to seek advice from stakeholder groups, collect evidence through trials, and learn from other cities are certainly hopeful, but it is hard to see how we can handle collecting e-scooters parked poorly everywhere in an eco-friendly way. The emergence of Tortoise explains that shared e-scooter services do not provide long-term profits for providers or help reduce their carbon footprint due to unintended consequences. And, after the introduction of shared e-scooters, an additional social experiment called self-driving e-scooters may follow since it could be ‘inevitable’. This makes me ask again. Why do we need a self-driving e-scooter? What should we learn from this?
