MPAT Tool Phase 3

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

THE DATA PLATFORM TO MANAGE FUTURE

MOBILITY ON CITY STREETS

Ai4Cities Mid-Point Check In

1
1
Agenda for Today

● Introduction and Background


● Concept of the MPAT
● Pilot Development and Evolution
● Artificial Intelligence
● Co2 Projections and Modelling
● Project and Work Planning

Vianova Confidential and Proprietary


Project Team

French start-up building the “Air Traffic Control for Urban


Dutch strategy & policy advisory firm, focusing on issues that affect
Mobility”: Cityscope our future, including sustainability, transportation, and urban
development

Frederic Robinet Alex Pazuchanics Ditti Osikovicz Francisco Macedo


Technical Product Development Project Management Policy Expert
Implementation

Jack Reilly
Back-End Engineering Kamen Todorov
Data Science

Vianova Confidential and Proprietary


Cityscope Core Architecture

Mobility fleets Cities


City data platform
Insights and
Mobility data
Intelligence
Shared micro-mobility Transport authorities

Ride-hailing Infrastructure managers


Monitoring and
Auditing Policy data
Micro- transit Public operator

Delivery

● Existing platform requires heavy human interaction


● Existing platform creates policy in a “dumb” way
● Existing platform monitors, but there is no baseline other than history

Vianova Confidential and Proprietary


The Mobility Policy Auto-Tuner (MPAT) in a Nutshell

● Recommend to cities the optimal distributions


of shared vehicles that would achieve the
highest CO2 reduction when balanced against
other priorities of the city

● Build the funnel from priority to policy- ease the


design and modification of shared mobility policy
requirements

● Continuously assess the effectiveness of


adopted policies in achieving CO2 reduction and
other policy goals

Vianova Confidential and Proprietary


Initial Pilot Approach- Implement and Monitor

-Goal of the Ai is to predict the baseline and compare against the reality of a new policy

-Difference between the reality and the projection is the effect of the policy

-Effect of the policy can then be quantified

Vianova Confidential and Proprietary


Revised Approach- Scenario Planning

-Cities are simply not able to move swiftly enough to implement policies and monitor results

-Takes time for innovation liaisons to find the appropriate piloting partners within city governments

-Cities have legal and political timelines for implementing policies

Vianova Confidential and Proprietary


Application of the MPAT
Tallinn Amsterdam
Develop the initial model of mode replacement Refine the initial model of mode replacement
From known data and industry best practice, what is the potential impact of From known data and industry best practice, what is the potential
one additional trip by walking, cycling, or shared mobility? impact of one additional trip by shared mobility (shared mopeds for
sure, shared cars if data is available)
Identify the policy toolkit to achieve changes to outcomes
How much impact would policy X have on increasing trips on a more
sustainable mode? Identify areas with the biggest potential area to target interventions
What are the range of available policies (scored on practicality, cost, impact, What geographies have the biggest mismatch between vehicle trips
etc.) and shared mobility usage?
How do we model the change in intensity of these policies?
Identify the policy toolkit to achieve changes to outcomes
(Potentially) Model the Co2 savings as a result of the change How much impact would policy X have on increasing trips on a more
How much Co2 would be saved if this policy was as effective as it could be? sustainable mode?
What are the range of available policies (scored on practicality, cost,
impact, etc.)

Model the Co2 effectiveness of the policy


How much Co2 would be saved if this policy was as effective as it
could be?
What is the projected difference between doing nothing and
implementing the policy?

Integrate with Cityscope to allow Creation & Monitoring of the policy

Vianova Confidential and Proprietary


Underlying Ai
-Goal is to predict 1) # of devices in an area each day (fleet size) and 2) # of trips commencing as a
result of that fleet size

-More complex to predict the change in # devices and # of trips as a result of a policy intervention

-Known challenges: effect of pricing, small geography of policies, normalization for weather

-Will ideally pilot projections for shared mobility, other active mobility, public transport (for Tallinn)

Vianova Confidential and Proprietary


Calculating potential CO2 impact (until 2025)
Existing market Modal shift Municipal
demand potential plans

Vianova Confidential and Proprietary


Does our Solution Create Emissions Savings?
-Our tool helps cities make smarter decisions
for the most Co2 impact

-Depending on the city’s readiness, the tool can


compare:

● A scenario with an intervention to a


scenario with no intervention
● The actual outcome of an intervention to
a scenario with no intervention

-Co2 savings are indirect but significant:


operating at the scale of the city

-Incredibly difficult to measure the direct Co2


impact of an intervention, but it can be modeled

- Co2 savings = Co2 Emissions of a “good” trip


vs. the Co2 Emissions of what that trip would
have otherwise been Vianova Confidential and Proprietary
Separate Components Converging in the Pilot
Done and Doing To Be Done

Data Science Creating a Model Projecting Demand and Usage Expand Reliability of Prediction
over longer time and smaller
geography

UX & Front End Wireframe the decision-making workflow Creating the “Scenario Planning”
Experience for the user

Methodology Developing a Modal Split substitution model for Apply methodology for Co2
Amsterdam and Tallinn emissions in the output of the tool

Policy Designing Scenarios for investigation based on Add ability to evaluate the
on-the-ground policy reality in Amsterdam and intensity of the policy intervention
Tallinn
Questions / Discussion

Vianova Confidential and Proprietary

You might also like