Oslo studiet –
Trafikmodellers anvendelse for nye teknologier
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
1
Søren Frost, COWI
Ruter
Existed since 2008:
› Public transport company
› Area: Oslo and Akershus fylkeskommune
› 1.3 million inhabitants
› Bus, tram, metro and boat
› Train is operated by NSB (YV)
› 371 million trips in 2017
Agenda at Ruter
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
3
Future role for Ruter:
› What is a public transport company role in the future?
› Is there a need for a public transport company?
› What will new technologies lead to?
› Self-driving vehicles
› Car sharing (shared ownership of the cars)
› Ride sharing (carpooling)
› Mobility as a Service (MaaS)
› …
Technological trends 1 (2017)
Project
• 3 consultant firms
• Literature Study
• Scenario development
Megatrends set the framework for future mobility in all scenarios:
• Technological development
• Urbanization
• Climate changes
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
5
Technological trends 2 (2018)
› Model calculations of the possibilities and consequences of ”extreme” future scenarios
› Inspired by ITF/OECD ”Lisbon studies”
› Assumption on behaviour/demand:
› One joined MaaS concept
› Starting point is car or/and bus- and tram trips
› Car respectively bus- and tram users act together as a group
› Demand is unchanged
› Calculated on the traffic for the whole Oslo region
› Full implementation is considered
SCENARIOS FOR FUTURE MOBILITY
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
7
Vehicle types and service level Capacity
4 persons Capacity
6 persons Capacity
20 persons
Model scenarios
- 4 base scenarios 1A, 1B, 2A and 2B
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
9
Model scenarios
- 2 additionally scenarios 3A and 3B
THE MODEL WE USED
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
11
› RTM23+ traffic model (EMME)
› Existing traffic model for Oslo
RTM 23+
Modelling system
VISUM MaaS Modeller
› VISUM traffic model
› Transfer of the RTM23+ traffic model
› Public transport (from headway to timetable)
› Demand model not imported, only matrices
› PTV MaaS Modeller
› New modelling tool to calculate on future technologies like Mobility as a Service
› Uses the VISUM traffic model as base for the calculations
Ride pooling as a mobility concept of the future!
The shared mobility algorithm addresses three core conditions:
› Minimise unserved trip requests
› Minimise the fleet size required
› Minimise the objective function (cost)
Simulating
Shared Mobility
(MaaS Modeller)
City focus
Passen focus ger
Operati onal focus
Holistic overview
The Objective Function
Passenger focus
Travel demand served
Waiting times
Detours
Travel distances
Travel times
Fare
Operational focus
Required vehicle fleet
Occupancy
Duration of the
trip Operating hours
Operating performance
Revenue
City Focus
Congestion relief / impacts
Environmental factors (emissions)
Safety
Urban realm possibilities / challenges
Sustainable
decisions
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
13
MaaS Model
Key Statistics
Scenario 2a
536,436 trip requests
56,000 vehicles
37,279,151 journey legs
Simulating Mobility as a Service
RESULTS
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
15
Best and worst case
t t
Fleet size reduction
-93%
Vehicle Vehicle
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
17
Empty vehicles
28% of the
kilometres is
empty vehicles
Fleet utilization
Vehicle operation distance increase from 12 kilometres to
about 150 kilometres
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
19
Level of service
Public passengers in bus/tram saves about 11 minutes Car users get extended
travel time by driving with ride sharing of
approx. 8 minutes Car users get extended
travel time by driving without ride sharing of
approx. 6 minutes
Network impact
Base scenario
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
21
Network impact
Scenario 1B
Network impacts flow volume vs base Volume / capacity ratio
(Period morning rush)
Network impact
Scenario 2A
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
23
Sensitivity analysis
Level of service
• Detour factor and waiting time
Larger reduction in vehicle kilometers and fleet size can be
achieved, but…
….it costs at the
service level. From 10 to 20 minutes
of accepted waiting time makes no
difference
Sensitivity analysis
Level of service
• Served passengers
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
25
Pick-up and drop-off locations
Scenario 1B
• Activity
Comparison with other
cities
Comparison of results with Lisbon and other studies
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE
28
Reduction in number of vehicles
Reduction in vehicle kilometres
26 AUGUST 2019
OSLO STUDIET - TRAFIKDAGE