Smart traffic

Smart traffic, or more broadly smart mobility, is one of the key initiatives of all smart city developments today. The challenges of the traffic largely include the same topics that drive also the development of the smart cities in general: fast urbanisation, mobility is-sues of the aging population, control and reduction of the climate change and pollution, mobility service development through innovative digitalisation, and discovery of sustain-able and efficient energy sources for the traffic (Hautala, Karvonen, Laitinen, Laurikko, Nylund, Pihlatie, Rantasila, & Tuominen, 2014).

In an assessment of urban transport, it is noted that the urbanisation and the related increase in road traffic will cause congestion and air pollution, simultaneously reducing the quality of life (Hajduk, 2016b). The EC has made a forecast that the freight transport will increase by 40 % and the passenger transport will increase by 34 % from 2016 to 2030. Thus, the EC has obliged the European cities to develop sustainable mobility strat-egies with the goal of improving passenger and freight traffic and reducing environmen-tal degradation in the cities. The proposed means to achieve this include the promotion of public transport as well as alternative forms of movement, like walking and cycling.

The coordination of timetables between different transportation means, the integration and creation of rhythmic timetables between train, tram, subway, and bus services, with properly planned interchange locations enable the creation of synchronised, multi-modal transport means. The development of intelligent transport systems allows the management of public and private traffic on the roads, including rail traffic, fleet, and cargo transport, and even information for the drivers about traffic congestion and the availability of parking spaces. Interestingly, the study also encourages the cities to invest in road construction, especially the modernisation of the ring roads and the exit routes from the city to the national roads are seen important. Still, for example, the city of Hel-sinki continues the controversial planning of converting its main exit routes into slower and narrower city boulevards (Lempinen, 2019).

The latest international research presents some interesting examples that widen the scope of the smart traffic concept to new areas of innovation. For example, the typical car-sharing services have so far used standard mass-produced cars. However, a design and manufacturing study in Bogotá, Colombia, attempts to create an electric super-com-pact vehicle uniquely for car-sharing purposes (Mendoza-Collazos, 2018). The design of the car is motivated by the desire to reduce the congestion with the small car size, the goal to preserve the user experience of a private car, and the wish to simultaneously improve the usability of a super-compact car.

Another study example presents how the computational power available today can be utilised in a traffic flow forecasting method that is based on a cascaded artificial neural network (CANN) (Zhang, S., Kang, Hong, Zhang, Z., Wang, & Li, 2018). The writers assume that this is the first study where CANN is used in traffic flow prediction. The developed system utilises open data and APIs to input and process weather information, map and route information, and traffic schedule, holiday, and behaviour information of the citi-zens to the system. The municipal road surveillance cameras provide pattern recognition information from the license plates to identify and timestamp the cars on the road. This information is fed into three artificial neural networks (ANNs): The long-term ANN cal-culates the periodicity of the traffic on a weekly lever, the medium-term ANN computes the daily periodicity and travel habits of the drivers, and the short-term ANN calculates the numeric variation trends of the flow of the traffic. The cascaded results of these three ANNs indicate promising performance and increased effectiveness in the traffic flow predictions compared to the more traditional prediction methods.

A prime example of a solution for the last-mile problem in multi-modal smart traffic ini-tiatives is proposed in the form of shared, short-term rental electronic scooters. These e-scooters promise sustainability, reduced environmental impacts, and the benefits of collaborative consumption as part of the burgeoning sharing economy. However, a re-cent study has noticed that the e-scooters may not necessarily reduce the environmental impacts, and potentially may increase the life cycle emissions in comparison to the trans-portation methods they replace (Hollingsworth, Copeland, & Johnson, 2019). More effi-cient collection of the e-scooters for charging, shorter distances of e-scooter distribution, and prolonged e-scooter lifetimes could reduce much of these negative effects.

Unfortunately, the recent news indicate that the lifetime of the e-scooters may often be calculated in days instead of years, that the sharing economy may leave the e-scooter collectors with low wages, while the riders increasingly find themselves injured by e-scooter accidents. A study by an online business publication reports from Louisville, Ken-tucky, that the average lifespan of an e-scooter is only about 29 days, the longest lifespan

observed was just 112 days, and about 4 % of the e-scooters disappear during their first day in service (Griswold, 2019). A Finnish newspaper reports that the e-scooter compa-nies may pay only one euro per a scooter for the collectors, with a possible freelancer agreement including inadequate employment terms and conditions (Harju & Nuuttila, 2019). Furthermore, the daily newspaper from Helsinki reports increasing amounts of injured e-scooter riders with fractured facial bones, even brain injuries, broken teeth and upper limb fractures requiring surgery (Kantola, 2019). A research from the United States confirms similar findings, with close to a fourfold increase in hospital admissions, with nearly a third of the patients having a head injury, due to e-scooter accidents between 2014 and 2018 (Namiri, Lui, Tangney, Allen, Cohen, & Breyer, 2020).