Evaluating the Environmental Impact of Electric Autonomous Vehicles through an Analysis of Travel Patterns

Voiceover

Presenter(s)(s)

Natalie Kelly, Adeline Hauck, Cassandra Starr

Presentation Number

701

Abstract or Description

The introduction of autonomous vehicles (AVs) will affect driver transportation patterns and Vehicle Miles Traveled (VMT). An anticipated rise in AV travel demand, thereby increasing VMT, will reduce vehicle lifetime and increase energy consumption and greenhouse gas (GHG) emissions. Given the potential for AV usage in the future and its undetermined environmental effects, this research aims to identify adoption and use factors contributing to society level VMT and the resulting implications for a life cycle assessment (LCA) of autonomous vehicles. Previous environmental assessments have not considered changes in travel patterns or effects of human behavior in analyzing the environmental impact of AV introduction. This research relies on (1) an initial survey gathering participants’ current travel behavior and willingness to adopt AVs, followed by (2) the utilization of a traffic simulation (VISSIM) to determine travel patterns for various locations in Michigan and AV repercussions. These outcomes will inform (3) a LCA of the use-stage of AVs to determine the net environmental impact in terms of GHG emissions and energy consumption. Analysis will be applied to Michigan residents and simulate mid-sized (SUV) fully electric AVs. Results will contribute to further understanding the behavioral implications of AVs and their related environmental impacts.

Mentor

Annick Anctil, Sharlissa Moore, and Mehrnaz Ghamami

Comments

Ebube Okpechukwu4 years ago

This presentation was laid out nicely and you all did a wonderful job explaining! AV is defiantly a area that needs more research on

• • 1 comment

Cassandra Starr4 years ago

Thank you Ebube!

Madison Janey4 years ago

Hi! Great presentation and very easy to follow. Question: Was financial security considered as a factor when studying the difference between the types of vehicles?

• • 1 comment

Natalie Kelly4 years ago

Thanks! Income was considered in the survey results and was significantly correlated with interest in owning an AV and moving further from work, so those who might be more financially secure would prefer to own rather than share an AV and are less interested in moving even if they had access to an AV for their commute.

Peter Savolainen4 years ago

Great project and video! A few questions: (1) Which of the key factors affecting travel patterns do you believe have the greatest (and least) amount of uncertainty...or where do you believe we have better (worse) confidence? (2) What are the underlying assumptions that led to the travel time reductions? For example, since these vehicles might travel at lower speeds, I would be curious as to the various mechanisms that would impact this aspect pf performance. (3) What are some of the key issues that you were not able to investigate given time/resource constraints and in what areas is future research warranted?

• • 3 comments

Adeline Hauck4 years ago

Great questions! (1) We found that many people have not thought about how an AV would affect their lives, so we have a great amount of uncertainty on how people would realistically change their travel patterns, as they might change their mind once AVs are closer to becoming commercial. We are especially uncertain about how frequent people may use their AVs as compared to their vehicles now.

Adeline Hauck4 years ago

(3) For the survey, we did not have time to interview as many people as we wanted and the survey participants tended towards the older population rather than the younger population. For the traffic simulation, there were some issues with incorporating AV platooning into running simulations. We were only able to successfully include AVs without platooning possible, but because platooning has a lot of potential to improve traffic flow, it would be great to explore more. We would definitely need more time to consider why the parameters used in the simulation were not working and resources that offer different calibration settings. Additionally, the LCA was constrained to the manufacturing and use phases, so disposal and material production could contribute to the overall environmental impact and possibly change our conclusion. Some future work could include looking at how social/travel patterns actually change once AVs are adopted, rather that relying on what individuals hypothetically expect themselves to do in the future. As more AVs are on the road, the possibility for platooning, or communication between the vehicles allowing for more efficient travel, could also reduce VMT. In the short term, the future of remote work could also be an important factor in how much individuals commute, since our results are based on explicitly asking participants about their pre-pandemic travel patterns.

Cassandra Starr4 years ago

(2) To get the values we did for travel times, calibration data for the Wiedemann 74 (urban traffic) and Wiedemann 99 (highway traffic) models were utilized from other literature - models utilized by PTV VISSIM for its simulations. The adjusted parameters were generally focused on accounting for the more aggressive driving behaviors we assume that AVs could safely do: decreased following distance, decreased standstill distance, increased number of observed vehicles on the road, deceleration values, or cooperative merging. All of these combined result in travel time reductions.

Eva Kassens-Noor4 years ago

Great work and wonderful presentation. If I understand your presentations, AVs, even if electric, would be worse for the environment given changes in travel behavior. How do you think we can incentivize travel behaviors to make AEVs more environmentally friendly?

• • 1 comment

Cassandra Starr4 years ago

During our literature review and research, we found several of sources suggesting ride-sharing as a good method to avoid empty Vehicle Miles Traveled (VMT) that a private AV might accumulate because it is only being used by a small number of people. Our group's thought was that if we can incentivize riders to share vehicles much more often, then that would be a good way to reduce the amount of energy being used by an AV to travel that is not actually being utilized by an individual. There's definitely still a lot of room for us to look into this - the biggest travel behavior that stood out seemed to be how we can expect increased VMT despite more efficiency from AVs because they increase accessibility for non-drivers, require less parking, etc.

Michael Hernandez Lamberty4 years ago

Really great project and presentation! What have you all considered to be the future work of the project? This is could in the short and / or long term

• • 2 comments

Removed4 years ago

This comment has been deleted.

Natalie Kelly4 years ago

Thanks! In the future, it would be interesting to see how travel patterns actually change for individuals with access to AVs, rather than how they predict they will change. Also, we could consider different mixes of increased/decreased VMT, adoption, and AEV/EV split in our LCA to reflect changing trends. The potential for relocation and changes in commute patterns will also heavily depend on how work from home continues after the pandemic.

Jessica Ullom-Minnich4 years ago

Great job! I am curious when conducting interviews to better understand the survey data, did you notice any correlations between survey respondents being more/less willing to adopt AVs and being willing to carry out a follow-up interview?

• • 1 comment

Natalie Kelly4 years ago

Yes! Most of the people we interviewed were really interested in or excited about AVs and had done some research on their uses. Other individuals who participated were incredibly concerned about safety risks, so the interview participants generally skewed towards people with stronger feeling about AVs either way.

Nrushad Joshi4 years ago

Great job everyone with the presentation and awesome poster.

• • 1 comment

Natalie Kelly4 years ago

Thank you Nrushad!

Katelyn Rousch4 years ago

Great Job! I really love how you've laid everything out, especially your final conclusion. It's really clear. Your QR code for references is very smart as well. At face value, AVs could make driving more appealing due to (a) not having to drive at all (b) being "environmentally friendly" because they are efficient. Did you factor the potential for the appeal of AVs to increase overall traffic into your model? If not, do you have any ideas on how you might look into that area?

• • 1 comment

Adeline Hauck4 years ago

In the survey we asked people how many cars they would replace in their garage if they were to purchase and AV. Using that situation, scenario 4 modeled the environmental impact considering an increase in total number of vehicles on the road (+4%). The increase in total vehicles also increases the net environmental impact. However, we did not consider other situations for total traffic in our model (such as people using AVs for long trips instead of planes, or people who did not previously drive starting to use and AV) and we did not factor in any situations of decrease in total vehicles in our model such as shared driving.

Benjamin Kessler4 years ago

Hi all - great work! I work with the URCA office, and I was wondering if you would be willing to have your research featured on our Instagram story this afternoon. If you have an Instagram account, I can tag you, too.

• • 1 comment

Adeline Hauck4 years ago

Yes! Thank you for reaching out. Feel free to tag us @adelin3_21 @natkel22 @caszcass

Chizaram Ugboh4 years ago

I really admire how you organized your poster. The QR code is such a great idea to save space. I know that you mentioned within your presentation that AVEs have more of an environmental impact than AVs in terms of changing travel behaviors. Could that finding be lessened if we changed the infrastructure around to accommodate these changing travel patterns?

• • 2 comments

Cassandra Starr4 years ago

Thanks for your question! I can imagine if we change infrastructure once we have AVs (maybe less parking, for example) that there might be room for improved traffic flow, more space. Spending less time on the road or using a car less will result in less energy consumption, which the traffic flow might help with. Energy consumption was the major aspect of AVs not being as environmentally friendly as we might have expected. I appreciate the question and it's definitely important to consider!

Natalie Kelly4 years ago

Also, one infrastructure change could be transitioning to more renewable energy sources and making charging stations that use renewable energy more accessible, so that the environmental impact of charging AVs is less significant.

Cynthia Liao4 years ago

Your video presentation is one of the best ones I've watched - so clear and everything flows so well!! Our project also looked at the environmental impact and we have indicated that as AVs allow people to perform tasks and work while commuting, they will have a better "tolerance" for long transit time and thus choose to live further away from where they work as housing near business centers are more expensive. This would then result in not only higher GHG emissions from commute but also from the construction of new houses and apartments, as well as the needed expansion of AV system to convert the newly developed regions. Just something to consider as the indirect environmental impact of AVs. Great job guys!!

• • 1 comment

Adeline Hauck4 years ago

Great points! We did consider increased vehicle miles traveled from people moving farther away, but we did not think about the need for more infrastructure, very interesting. Thank you for checking out our poster!