Welcome to our journey towards a greener path of mobility. The push to reduce CO2 emissions is clear, but figuring out our own carbon footprint - and how to make a difference - can be complicated. In this blog post, we explore the critical issue of CO2 emissions from vehicles, using real-world data from Astara Connect. We delve into the factors affecting emissions and present actionable insights, including shared mobility options and electric vehicles as viable steps to minimise our environmental footprint.
CO2 emissions from vehicles
When your car operates, it releases carbon dioxide (CO2) from burning fossil fuels like petrol or diesel. During this process, the carbon atoms from the fuel combine with the oxygen in the air, emitting CO2 into the atmosphere through the car’s exhaust system. The amount of CO2 released depends on several factors, including engine type and size, vehicle weight, fuel efficiency, and driving habits.

Typically, vehicles with larger engines or carrying heavier loads emit more CO2. In contrast, cars with smaller, fuel-efficient engines emit less. Furthermore, driving habits like speeding, prolonged idling, and aggressive acceleration can increase CO2 emissions. This last factor is key because it gives us the power to impact our emissions through our actions regardless of the car we drive (of course lower consumption engines will make a greater difference).
Revealing Our Data: Description and Sources
Welcome! Astara Connect is a pioneering low-code mobility platform that turns your operational vehicle data into valuable insights. By generating and analysing data, Astara Connect helps its users to spot patterns to optimize their fleet mobility. The dataset we're discussing in this blog post gives you a glimpse of what we do with the rich data from connected vehicles. This rich dataset not only speaks volumes about vehicle usage but also sets the foundation for our CO2 emission estimates.
Our dataset includes recorded trips from 252 connected vehicle over a 45-day period, spanning diverse locations in Chile. These vehicles, all shiny and new from 2023, feature eight different models from the respected manufacturers, Mitsubishi and JMC motors.

Recorded trips summed up to 22,700, spanning an impressive distance of 172,000 kilometers collectively. On average, each vehicle covered approximately 368 kilometers. Over the 45 days, each vehicle made an average of 91 trips, spending around 4785 hours on the road. Check out the graph below for a neat representation of these key trip characteristics.

In the first graph we see that most cars traveled short distances (from 0 to 1000 km) that means that shorter trips are the most prevalent (second graph) and the third one shows that a handful of cars had the most trips.
You'll see from the next graph that the number of trips ramped up over the 45-day period. This shows us that car usage has been consistently growing over the period we analyzed.

Unlocking the secrets of WLTP values for CO2 emissions estimation!
So picture this: after deep-diving into our dataset and understanding our vehicles' behaviour and trip characteristics, Our mission? To crack the code on CO2 emissions. And guess what? We hit the jackpot with the WLTP CO2 values! But hold your horses, these values aren’t just numbers: they’re the key to unlocking the mystery of CO2 emissions, using data from connected vehicles, such us the distance traveled. Sounds cool, right?
Now, let me introduce you to WLTP - The Worldwide Harmonized Light Vehicle Procedure. This is a global standard that helps us measure the fuel consumption, range, CO2, and pollutants emissions (like NOx and particulate matter) of all kinds of cars, from traditional to hybrid to full electric.

Why the fuss about WLTP, you wonder? Well, it’s here to bridge the gap between lab tests and real-life driving. It’s a step up from its predecessor, the New European Driving Cycle (NEDC), because it includes higher speeds, diverse driving scenarios and more realistic accelerations and decelerations. The goal here is to paint a truer picture of a vehicle’s performance.
You can find WLTP data in your car’s manual or on the manufacturer’s website. Even regulatory agencies like the EEA chip in by publishing WLTP values. Now, here is the kicker: WLTP measures CO2 emissions in grams per kilometer (g/km), telling you how much carbon dioxide a vehicle releases for every kilometer it travels under test conditions.
But wait, there's more. You can even estimate total emissions for a given distance by simply multiplying the CO2 value of your car by the distance of any trip you've taken.
In our data adventure, we used the make and model of each vehicle to fetch CO2 emissions values for each car model from the manufacturer’s site. And guess what? The graph below shows the WLTP CO2 values for the vehicles in our dataset. Surprise, surprise - the diesel darlings, Touring minibus 2.8 and L200 Dakar 2.3 top the CO2 emissions charts per kilometer. No shocker there, since diesel fuel and big engines tend to emit more CO2. It’s proof that engine type and size play a role in emissions.

Ready to explore more?, Buckle up, because this data journey is just getting started! 🚗💨
Cracking the Code: Estimating CO2 Emissions
So, after crunching the numbers, we’ve got the CO2 emissions for each trip made by every vehicle in our dataset. Drumroll, please… On average, each vehicle released 120 kilograms of CO2 over the 45-day period. But, here is where it gets interesting - the emissions extremes. We’re talking the highest amount recorded by a single vehicle was 840 kilograms, while the lowest? A mere 1 kilogram.
Now, picture this: the difference between the top CO2 emitters and the champs at minimizing emissions is downright dramatic. The top ten emitters released between 495 to 840 kilograms of CO2. On the flip side, the CO2 minimizers, emitted as little as 1 to 1.15 kilograms of CO2. Talk about a showdown between the heavyweights and the lightweights, right?

But wait, there’s more! You’re probably wondering how our CO2 emissions stack up against an average petrol-powered ride? Let’s break it down. Consider this average petrol-powered car with a fuel efficiency of 10.6 kilometers per liter and an average mileage of 1600 kilometers. Over that same 45-day stretch, it would release a total of 529 kilograms of CO2. Now, compare that with our’s fleet average CO2 emissions of 120 kilograms. Mind blowing, isn’t it? And get this - half or our vehicles emitted less than 100 kilograms of CO2. Talk about doing our bit for the planet!
Deciphering CO2 emissions: What it really means
Let’s put these big CO2 emissions into perspective, shall we? Over 45 days, our dataset showed total emissions of 30,408 kilograms of CO2, which is a whooping 30.48 tons. But what does that really mean in everyday terms?

Imagine this, the total CO2 emissions from the 252 vehicles are enough to charge 3.4 million smartphones! Or if you’re more of a binge-watcher, it is equivalent to the energy needed to stream videos for a mind blowing 60,800 hours.
If you’re more of a foodie, here’s another way to look at it. The CO2 emissions from our vehicles are the same as cooking up 4,200 meat-based meals. That’s a lot of cooking! And if you’re more of a coffee enthusiast, we’ve got you covered there too. Our dataset emissions could brew up 268,000 cups of coffee.
Unveiling the influence of vehicle traits on CO2 emissions:
Okay, by now we all know that the trip distance has a big impact on CO2 emissions. But have you ever wondered about the role of a vehicles’ fuel type in this?
We’ve cooked up something special in our article to explore this relationship. We whipped up a little variable called the “petrol model”, which basically tell us if a vehicle runs on petrol (gets a value of one) or diesel (gets a value of zero).
Now check out the box-plot below. It reveals that diesel vehicles generally puff out more CO2 than their petrol counterparts. But here’s the twist: when you look at the big picture, petrol cars, which are more common in our dataset, are actually the biggest contributors to CO2 emissions.

We’re about to take an exciting journey into the world of CO2 emissions from our vehicles, using the magic of a clustering analysis. Our trusted tool for this mission? the k-means clustering algorithm! It is a popular unsupervised machine learning technique that groups data based on how alike they are. With k-means clustering, we’ve sorted vehicles into three groups. They’re based on their CO2 emissions, number of trips and the duration of trips.

So, what did we find out from this clustering process? Here’s the scoop:
- Cluster 0 - Meet the Minimisers: Picture this group as te eco-warriors of our fleet. They’re all about the lowest emissions, fewer trips and shortest duration. With 141 vehicles in this gang, they might be the largest crew, but they churn out the lowest emissions, totalling just 3,778 kilograms of CO2. And guess what? Their go-to ride is the Xpander Cross 1.5.
- Cluster 1 - Say hello to the Maximisers: Now, these guys are the heavy hitters. They are all about high emissions, frequent trips and long trips. Sure, they might be the smallest group with just 27 vehicles, but they pack a punch, accounting for the 43% of the total CO2 emissions in our dataset. And their ride of choice? The L200 Dakar 2.3!
- Cluster 2 - Meet the Average Emitters: Last but least, we’ve got the middle of the road crew. This groups is all about balance, with moderate values across emissions, trips and duration. Even though they're average in individual CO2 emissions, they contributed a whopping total of 13,616 kilograms of CO2. Just like cluster 1, they’ve got a spot for the L200 Dakar 2.3.

Let’s take action: Reducing your CO2 footprint
So far, we’ve embarked on a journey together, learning how to calculate your cars’ CO2 emissions and understanding the carbon footprint of vehicle users. But now it’s time to roll up our sleeves and make a difference. Whether you’re already acing the eco-game or need a little nudge in the right direction, we’ve got some handy strategies to help you cut down your carbon footprint.
- Public Transportation, Shared Mobility and Micro Mobility
Chances are, you’re already aware of public transportation in your area. But, we get it - sometimes it might not seem as convenient or flexible as having your own ride. However, shared mobility options are always evolving and can be game-changer when it comes to slashing your emissions. So, next time you’re planning a trip: Why not take a moment to explore all your options? Compare total time including parking and total costs, you might find something that suits you better.
- Electric Mobility
Now, this one is a bit more complex as it requieres some major changes. But fear not! We’ve put together a compatibility study to help you figure out if an electric vehicle is the right move for you. Trust us- It’s easier than you think!

The graph you see above shows the main factors to think about when considering a switch from combustion engines to electric vehicles. Now the first parameter -Costs- is pretty personal and we won’t go into it here. Instead we’re going to focus on the other factors using trips’ data.
Next up in the graph is the Range factor. This refers to how far an EV can go before needing a recharge. In the past, the big question was whether an EV could keep up with your travel habits. But nowadays, with charging infrastructure and EVs that can go further, the real question now is how much tweaking you’d need to do your habits to make the switch.
To figure out if an EV can handle your travel needs, we’re going to look at some trip data. Let’s take a peek at the trips of our Chilean user over the 45-day stretch. The graph below summarizes the trip statistics for one particular user. This user makes about 100 trips a month, which is pretty average if you think about the usual home-commute-home or home-commute-household_responsibility-home routine, making up 90 or 100 trips a month, plus a few extra on the weekend. This user mostly takes short trips (between 0 and 10 kms), with fewer than 6 trips going over 40 kms. These are distances that an EV can handle with ease.

Another thing to think about is trip frequency. The chart below shows the trip frequency for the first 15 days of February, with weekdays in coral colour and weekends in blue. Looks like our user is pretty active, with around 7 trips per day on weekdays. But here's the kicker – most of these trips are short-haul, which means they're perfect for an EV.

But it's also important to think about where these trips are taking place. Are they in urban, suburban, or rural areas
So, let’s talk about territory. We’ve mapped out our user’s trips in the urban sprawl of “Gran Santiago”. And guest what? All those blue lines showing where most of the trips took place? They’re firmly within the city limits. Now, why does that matter? Well, densely populated areas like this tend to have more charging stations – a win-win for EV owners.

The area where most user trips were clustered was surrounded by stations. ¡Good news!
The black dots in the following image show a sample of the total charging stations found in Santiago.

There you have it, folks – your roadmap to cutting down on CO2 emissions and doing your part for the planet. Stay tuned for more tips and tricks on living your best (and greenest) life!