Introduction

My name is Sara Iglesias.

Who I Am and What I’ll Cover

I'm part of the team of Liftango, the European partnership of lifts. I'm going to talk about Liftango and AI and shared mobility and how it comes together.

Framing the Core Question

So tonight I want to address this question.

Smart Mobility and AI vs. Traditional Transport

How do smart mobility and AI outperform traditional transportation solutions?

This is transportation we are all connected with.

Why Transportation Needs Change

I wanted to see it with the lens of this is a complex logistic problem that not only big cities are facing now, but it's an environmental problem.

A Personal and Professional Background

So for that, as it is a tricky question, I want to give you an overview and a background of who are you listening to tonight.

So I'm a sustainability advocate. I have done many post grades in sustainability in the corporate level. I was part of the committee of sustainability in the previous company. I was working at there was the biggest transportation company in Spain

I'm a mobility expert. I am all my background. It's in in this sector

and actually I'm coming from a company family that it's built by transportation Roots, this is the the company of my grandfather in Spain Spain.

I'm coming from the north of Spain, Galicia, and this was the shadow line that connected the airports from Portugal to Galicia.

From AI Enthusiast to Practitioner

Obviously I'm here, as Sergi said, I come here to a lot of the Mindstone events, I'm an AI enthusiast, and I've been studying AI since the beginning of my professional and academic career.

I study computational linguistics, so it was basically the beginning of AI and how to transform natural language and process process it to machine learning, building sentiment analysis tools for example.

So my goal since I left uni is to bring that idea, academic ideas to the real world and to make a solution for them.

Current Pain Points in Mobility

So our transport systems are stuck in the past, it's not the most updated sector let's say.

So there is a lot of work on education to do with the operators, with the cities, with the transit agents.

uh you can the pain points here and the reason i'm talking about this are the congestions the traffic jams we all have felt them i i i'm sure you all know about the traffic lights here in geneva and that's the reason my car is happier in the garage um the emissions um i would like

Commuting, Emissions, and Equity

to know how many and i'm not going to judge you but how many people here use the car to go to work Can you raise your hand? Shame. No, I'm joking.

1But the 68 % of the missions of the Scope 3 are from people commuting. Scope 3 comes from companies that are not manufacturers, are like finance, software. So they don't have directed missions like Scope 1, Scope 2. But the Scope 3, it's mainly from people commuting. I'm not trying to make you feel guilty, but this is what I'll do, taking your card.

And then obviously transparent equality, this means underserved areas like communities that don't have very good connections, they don't have access to their jobs, hospitals, universities, if they don't have a card, and they are isolated by that.

Forces Driving Transformation

So which are the drivers forcing the change now?

Regulatory Pressure and Decarbonization Mandates

Now, the regulatory push in the EU, now we have the CSRD, Corporate Stability Reporting Directive, that is creating this pressure. We have the visions that I was talking about before that needs to be reported. This is becoming a must for companies now, and it's a massive requirement.

And we have the carbonization mandate, so the COP29, that is also a demand from governments now, that it's not like you have to have a plan you have to have a working plan and show that it's actually working and these lead us to the main AI issue and

The Data Gap: You Can’t Fix What You Can’t See

the solution of course for it that we can fix what you cannot see that means there are huge data gaps in terms of data of passengers and transportation especially in these underserved areas that I was talking about so So without no data, no insights, it's difficult to recreate what people want and what people need for those areas that have no existent transportation.

And obviously, it's a high risk and high cost to deploy a huge transport line routes without any idea if people want that or not.

Liftango’s Mission and Footprint

So for that, Liftango, the transportation company, software mobility company that I'm working for has a goal that is to prevent one gigaton of CO2 entering the atmosphere by optimizing this public transportation and commuting people around.

run.

Who We Are and Where We Operate

The company was founded in Australia in 2017. It has five global offices, more than 150 networks from public, corporate and community services.

Just to show you that even though it's a kind of new company, it has built a lot of partners and clients around the world.

You can see some very well -known that that are actually here also around the Alps region, like Lufthansa, Expedia, Tesla, Amazon, a lot of universities.

Just a quick overview also on how much it has been growing in the public sector, but also in the corporate transport sector and all the investors that has been trying to support a company that's looking for a positive impact.

The Platform: How It Works

So how does the platform pass it? it's a hub for a total mobility platform that means it has the public transport site the corporate transport with each of the layers that we will see and the

AI, Data, and Applications Across Sectors

data that this is the part that is powered by AI and machine learning so their transformation the real world for the platform and how it's applied in the

public sector for example we help cities and government to optimize the existing routes that they have.

We do it with last mile for example for these rural areas that I was talking about that it's hard to access for them.

Community transfer like elderly or handicapped people and school transport.

Corporate Use Cases and Decarbonization

Then in the corporate side tackling this scope 3 CO2 emissions that I was mentioning before we do it for campus, for business parks, airports, university and then

And the decarbonization goal, it's also by implementing AB electric vehicles and hydrogen. So the AI tracks the charging time and takes that in account for the optimization of the route, for example.

We do that with Facebook, with National Express in the UK, and MTA in New York.

Demand-Responsive Transport, Explained

So how does the demand -responsive tool work?

Predict, Optimize, Learn: The AI Brain

it works first on the band prediction engine so the no no data problem that i was talking about it generates that data from the scratch we will see after it has the dynamic route optimization

with that data then chooses the best scenario depending on what you're looking for the fleet management intelligence that means if the car if the vehicle needs maintenance or not it takes that into account and remove it or put it apart in the scheduling system and of

course the platform the user experience system if you're a passenger the app if you're a driver you're up to take care of the passengers so since we are

talking about AI this is how the brain works predict optimize learn so as I was saying, it predicts the demand with a cartographer that I will show you. It optimizes that data and depending on what you're looking for and it learns.

So the machine is not the same the first day as the day 100 because it learns by the movement patterns, the delays of the driver, the events in the road that might happen.

Service Designs and Deployment Models

And so just very fast, I'm going to give you a general view of how it looks like in service designs.

Door-to-Door, Hub-to-Hub, and Hybrid Approaches

So for example, door -to -door would be for a low -density area that it has a lot of kilometers to cover, that a fixed route wouldn't be the best tool, the best option to put there.

So it can cover a lot of the areas, but in an on -demand weight.

Then the hub -to -hub, that would be the micro feeder for example, like connecting a train station with a pier and having that last mile also.

And then we have also the fixed route skips, stops and zone -to -zone, that would be a hybrid type of light that can be a fixed route that can go to the last mile and deviate it from the fixed route if needed. it.

Simulation and Optimization Tools

So this is the transport simulator. This would be the part of where AI and machine learning take place and how to optimize this real -world assessment and see how with the data they provide us.

For example, X amount of vehicles that we want to deploy, we have this amount of employees if that would be the case of a corporate but if it would be the case of a government or a city hall trying to implement this it would be more like with employment centers or data that

Cartographer and Scenario Testing

we have from other sources so this how is this is how the cartographer looks like you can see here what you can select in terms of flexibility in the band and how it takes the data from that gap just creating a radio from it and this would be what the

simulation looks like after so you can see the metrics here in the success rate rate in this case is 94 .7 you can see the waiting time and where's the waiting time yeah two minutes for example and well a lot of metrics that can help us or help the partner to decide if this is scenario it's what they are looking for or not yeah we can run hundreds

of scenarios as you can see here too and the good part about this is that the partner if it's a corporation of its government it can decide what they are looking for are they looking for an equal the most eco -friendly scenario are they looking for being a boost for the passenger the passenger would take it more because it's very frequent or are you are you looking for a passenger comfort like spending less time possible in the bus for example so

Inside the Routing Algorithm

to make it understandable or how we look how it looks and how the AI chooses the

best route I don't this that it's quite simple explanation so when you have a series of requests how does the AI select which is the best route and how to select it. So, how does it differ from the route A to the route B?

Cost Functions and Fairness

It creates a score that is called the cost, and that cost would help us, would help the tool to decide which one is the best route. Obviously, less cost, the better.

So, here is a simplified way of explaining the algorithm and how it works.

It would be the route cost, it values the cost efficiency for the route, so fewer kilometers and how the shorter time for the passenger to be in that route, and then the passenger cost that

it penalizes, for example, a client that you cannot keep, a passenger that you cannot pick up or unassign passengers. So obviously this is a very logical example.

If the person that is being picked up in the last mile is going in the middle, the route wouldn't be efficient. So this is a calculation of you have to sacrifice the person who is the farthest because that person is going to spend more time in the route than the other people.

but at the end of the day what it tries to do is create a sense of fairness so it creates a total travel distance between that measure of the time deficiency of the route and the efficiency of the passenger itself and that results in reduced travel time and passenger experience and just to show

Real-World Impact

you in how this is applying the real work I selected some examples I have too many i had to choose some of them that were valuable but here for example you you can see

Global Case Studies

other other is a real estate provider in the emirates they have a lot of different businesses and we do also the school and the hotel and the school but for example the staff transportation has been optimized with a fleet of the missection of reduction of 30 % so buses are taken out of the road because they are applying the the tool to optimize

the regular lines that they had for the employees you can see here in the in the gap in the picture how the commuters are are being deployed I'm not going to stop in all of them some of them for example this is the Brazil one in some

paulo this i especially want to mention because this is the largest deployment of drt in the world so it means 6 680 vehicles that are now transitioning to a demand mode and it implies a lot of types of transportation like school buses medical care of paratransit and and general public transportation.

So this is how a city it's implementing new technology that is allowing them to reduce by 30 % the fleet.

Another ones that I'm not gonna mention, but they're going to rest it to you.

For example, when I was talking about the lap, the data optimization tool without being deployed, This is a study we are doing for the Dubai airport and see how the shuttles can be optimized.

This is another project that is already implemented, ring and ride in Birmingham.

And this is, I just left it for fun because I asked the technical team, the mobilization team, if I could share the data of one of the biggest clients that we had. and then he said sure like you can and started to discover everything with squares so this is what I got but still you can see how many co2 are being safe

which is quite nice and this is also one of the the bright treasures of the

AT Local, New Zealand: Electrified DRT at Scale

company the AT local deployment the project so AT local it's in New Zealand and this has been the alternative of our route that is the 371 and has managed to convert a lot of people that were not using the public transportation into actually using it it's fully electric so So this is already used in the electric vehicle platform part and it's very simple for the

rider. Like they just have an app, they book the ride, they go to the pickup point and they just pay with the local transit app.

So some figures, the coverage before using these 80 local operations, it was 16 ,817 passengers, which now the population has become 23 ,280. So this is a 38 % additional population that is using and now this type of services just because it's covering more areas in a more dynamic way and not in a fixed way.

Behavior Change and Mode Shift

Yeah a little bit too much and this for me is one of the most important slides because it shows how it's changing behavior of people so we asked the riders in New Zealand what what they thought about using AT -Local and how much has they changed in their behavior.

And we saw that 33 % of people would just take, I guess, the original route, the 371, but 60, 60 or is it 67? Yeah, 60, how much was it? 67 percent would take the the private car so that's a huge conversion of people that would actually are not taking cars anymore to work worth work to in their life and you know uh so

Conclusion

to wrap up what i'm trying to say is that um ai is doing a lot of good for this we are actually taking cars out of the road we are reducing the vehicle miles we are lowering emissions per trip per passenger and we are supporting green infrastructure.

I mean so the EV and hydrogen vehicles that are getting implemented and optimized by this platform.

Why AI and Smart Mobility Outperform Tradition

And to come back to the previous question of how AI and shared mobility, and smart mobility sorry, it's outperforming traditional solutions.

Well it is because it's predicting, it's optimizing and it's learning from all the all the data that has been gathering and it has proved that it fundamentally changed the behavior of people.

So thank you very much for listening.