International Benchmark on Experimentations with Autonomous Shuttles for Collective Transport

Mots-clés:

Résumé:

1. Purpose
With the rapid growth of urban population, urban mobility has become a key factor affecting citizens’ well-being and life-quality, specially in large centers where mobility solutions are urgent (United Nations, 2018; Melis et al., 2015). Recent innovations in technology and digitalization have had a great impact on designing sustainable mobility concepts: on-demand mobility services, autonomous driving, dynamic pricing algorithms and vehicle electrification will change the way people experience mobility (Alazzawi et al., 2018).
Autonomous Vehicles are expected to fundamentally change – for the better – urban mobility in cities (Lang et al., 2016; Mutz et al., 2016). However, fleets of AVs will not likely be seen on the roads right away (Mira-Bonardel & Attias, 2018), for the authors, it is likely that fully AVs may firstly be authorized for collective transportation, thus offering a solution for larger cities that struggle to provide adequate public transport solutions. Ergo, the emergence of Autonomous Shuttles for Collective Transport (ASCTs) promise to harness connected automated vehicles to catalyze Mobility-as-a-Service (MaaS) schemes in such crammed areas where mobility solutions are urgent.
Over the past years, a significant group of entrants have been testing Autonomous Shuttles for Collective Transport (ASCTs), however, there is still much to learn about these vehicles’ operations, given that their implementation scope is still scarce, non-structured and pulverized. Therefore, the purpose of this study was to perform a worldwide benchmark on experimentations with ASCTs.

2. Research design
The adopted research design was characterized as qualitative and quantitative of exploratory and descriptive nature and took as a starting point for the experimentations’ query the studies of Charlet and Chaufrein (2017), Hottento, Neines and Pinckaers (2015) and Mira-Bonnardel and Attias (2018).
By using the snowball sampling technique (Penrod et al., 2003), data was collected from September 3rd to September 18th 2018 on both academic and grey literature (both on structured and non-structured data). Saturation criteria (Fontanella, Ricas & Turato, 2008) was used as a stopping point for data collection. The research corpus consisted on 92 ASCTs’ experimentations worldwide. Next, data was structured and analyzed via descriptive qualitative analysis (Kim, Sefick & Bradway, 2016) and descriptive statistics.
It is worth highlighting that it was not the scope to consider the experimentations with AVs (up to five occupants) such as the ones provided by Waymo, Uber and, Lyft; neither did we consider experiments regarding autonomous vans/trucks for cargo/freight deliveries.

3. Main findings
From the 92 experimentations, 50 had already been finished, 31 were ongoing and 11 were yet to start. These 92 projects unfold in 78 cities spread over 32 countries enabled by 20 different manufacturers. Results show an European lead on both number of experimentations and manufacturers, with highlights to French startups Navya and EasyMile – which are the global leaders regarding ASCTs’ manufacturing and deployment.
Most experiments (96.55%) were fit within the Business-to-Consumer (B2C) business model, being those either trials (69.23%) or showcases (20.88%), mainly offered free of charge to commuters; whereas paid regular services were a minority – 9.89%. Business-to-Business (B2B) models composed the remaining minority of 3.45% of the sample – here, shuttle manufacturer choose to sell its products to private firms or instead, to provide transportation services to such companies. Peer-to-Peer (P2P) business models have not been identified in the sampled ASCTs – in such models, vehicles’ ownership are normally in the hands of ordinary citizens (peers) which by the use ridehaling platform-companies provide services to other citizens (commuters) (Macmurdo, 2015).
Regular-line transport systems were the prevailing operational mode adopted (91.21%), meanwhile, demand-responsive transport systems and mixed services were also observed, however incipient (8.79%). With that, as more countries and cities begin to allow testing and circulation of AVs on their roads and highways, the percentage of on-demand autonomous mobility is likely to increase, since the major value proposition claimed by ASCTs’ manufacturers is to facilitate the fist- and last-mile commute.
Eight main typologies of uses able to fulfill both B2B and B2C purposes were identified, being either focused on solving such first- and last-mile issues and/or microtransit commute. They were divided among closed traffic and mixed traffic conditions and also between regular-line transport and demand-responsive transport. Nine key-performance indicators were selected and divided into economic – and user-centered. At last, the main common stakeholders among all experimentations were identified, as well as how different forms of value (financial; usage; research; data) are created and distributed among them in order to promote a sustained growth and evolution of the ecosystem.

4. Practical implications
By attempting to to shed some light on worldwide experimentations with ASCT, the present study sought to highlight the most relevant shuttle manufacturers as well as countries and cities with most deployments; propose a relevant typology of uses for ASCTs (evidencing the nature of the deployed experimentations, revue models, prevailing business models, their classification within urban transport and, relevant key-performance indicators); and identify the main stakeholders involved and how value is created and distributed among them.
ASCTs are likely to be the gateway to mainstream use of AVs (Mira-Bonnardel & Attias, 2018). The raise of such shuttles promise to harness connected automated vehicles to enable Mobility-as-a-Service schemes (Harris, 2018), since their main goal is fulfillment of the first- and lastmile requirements as well as microtransit for city centers, central business districts, campuses, airports, shopping malls, hospitals, etc.
Even though there is still much to learn about the operation of these vehicles from both policy and regulation perspective as well as regarding businesses models and consumer acceptance (Harris, 2018), a significant group of entrants have been testing such automated mobility solutions for collective transport (Mira-Bonnardel & Attias, 2018; Clausen, 2017) with highlights to the French-based companies Navya and EasyMile.
At last, ASCTs are imbedded in a business ecosystem where growth and evolution depends on synergy and value flows among the stakeholders, even though, further studies are needed to empirically evaluate such dynamics.