ioki – mobility on demand #fmcdd19

I have to admit that before the Future Mobility Camp I neither knew the start-up ioki nor did I know that it belonged to Deutsche Bahn. When I learned about the latter during the session planning, I noticed that I was immediately struck by a certain basic scepticism. A start-up that belongs to the German railway and wants to turn public transport upside down? Can this really go well or won’t it rather be a marketing gag? Can a steamboat as big, old and in urgent need of modernization as the Deutsche Bahn really build up and maintain a young and dynamic start-up?

ioki’s analysis of 2019 and imagination of 2025

Apparently Deutsche Bahn can do this. ioki’s pitch was a bribe for its striking independence from Deutsche Bahn – it was not until the round of questions that the auditorium came to talk about the mother-daughter relationship. Otherwise, ioki presented itself as an independent start-up that wants to improve public transport (revolutionizing would probably be too far-fetched). While in 2019 the average car driver still searches 118 hours a year for a parking space and spends 75% of his time alone in the car, ioki wants to increase the share of shared and inexpensive trips in Germany by 2025. These are best done in autonomous vehicles.

ioki combines motorised individual transport and public transport for individual public transport on demand

While ioki is of the opinion that start-ups such as Uber or Lyft displace short-distance, cycle and pedestrian traffic, ioki aims at an intelligently networked public transport system that is supposed to enable flexible mobility without owning a car. Motorised individual transport will be combined with public transport to create individual public transport on demand. This is made possible by a platform on which customers can enter their travel preferences. Once the request has been successfully met, customers are picked up by a publicly operated shuttle at a public stop and taken to their desired stop with other customers collected en route. This, of course, takes place route-optimised, in coordination with the cycle times of other public means of transport, such as suburban and underground trains, and purely electrically. The concept is currently being tested in Hamburg in the districts of Lurup and Osdorf in cooperation with the Hamburger Verkehrsverbund (HVV) – enclosed are the most important key data for Hamburg:

  • 25 purely electric “London-Cabs” with a range of approx. 200km.
  • 110 offered public stops.
  • 200 passengers per day.
  • A shuttle covers an average distance of 500km per day.
  • The average occupancy rate is 1.7 (compared to 1.1 in normal traffic).
  • The shuttles are continuously in motion and follow an “intelligent” route. Similar to predictive policing, which predicts the next criminal offence, an algorithm is used to estimate where the next request is most likely to be made.

So much for ioki’s publicly available business concept. Much more exciting for me, however, was the procedure for selecting the Hamburg Lurup and Osdorf test field. In addition to the OnDemand platform, ioki has developed a “Mobility Simulator” that can be used to identify potentially suitable areas for the shuttle service – in other words, the Mobility Simulator seems to be the heart of ioki’s business concept. In concrete terms, a microscopic image of the areas to be investigated was created over a development period of 1.5 years, in which it was simulated for each individual inhabitant how they move through the public space. Due to the agent-based model, it is even taken into account whether it is a work, leisure or shopping trip of the individual agent. This detailed simulation was made possible by the fusion of different data sources, e.g:

  • Social demographic and geographic data (e.g. from national statistics offices)
  • Route diaries from mobility studies (e.g. Mobilit├Ąt in Deutschland)
  • Source and target matrices from mobile radio data

I find ioki’s use of mobile data from “a large German network operator” the most interesting. What could we accomplish in the field of active vehicle safety if we could use driving profiles from mobile phone data for driving and driver behavior analysis? On my request, ioki had to buy the mobile data on the one hand and on the other hand was subject to high data protection requirements. The mobile phone data used cannot be traced back to a single person, since ioki is only allowed and able to evaluate the data in relation to the municipal boundaries. This means that no driving behaviour can be reconstructed within a “municipality” – the macroscopic level of observation thus achieved is really very high.

All in all, I was really impressed by ioki’s presentation, but it remains to be seen how ioki’s model will perform. In any case, it is clear that this form of mobility will not pay off for the regional transport companies without a public subsidy – it is and will remain a local public transport system. On the other hand, it will be exciting to see up to which city size ioki can offer its services economically. For such a complex and detailed simulation not only a corresponding budget is necessary, but also a current and detailed data basis. I am curious!

PS: Whoever compares the basic idea of ioki’s shuttle on demand with the old-fashioned and often smiled at “call/collection bus” is a rogue ­čÖé