David von Dollen, former Head of AI at Volkswagen and now the VP of Data Science at GridMatrix shared his insights into AI for intelligent infrastructure and transportation at Worldwide AI Webinar.
David identified some challenges and opportunities that he had witnessed in the intelligent transportation space with AI enablement for the audience after walking us through the history of AI. For a complete history walkthrough, please check his whole keynote on our website
Keep reading for the highlights of the second part of his talk.
Opportunities and challenges of intelligent transportation
According to Statista, the urban population has been growing over the last 70 years. Clearly, cities are getting more densely populated, congested and polluted.
A 2018 study reported that around 75% of transport emissions came from road vehicles. It’s safe to say that automobile is responsible for quite a few global issues, including air pollution and deaths caused by transport accidents.
Though the automotive industry has made some great efforts, such as producing hybrid and electric vehicles to reduce carbon emissions, there is still much to be done.
Hence, David put forth the question:
“What can we do as AI practitioners to help with this problem?”
He proceeded to cite an article showing Toyota’s introduction of vehicle-to-infrastructure communications (V2I) in 2018 and stated that cities are generating a lot of data using different devices such as LiDAR, cameras, and induction loops to understand how our infrastructure is being used.
David claimed that would be a huge opportunity towards improving AI or implementing AI to allow for automated feature extraction and to allow human operators to extract more knowledge from and reduce the complexity of these systems.
Additionally, companies like Google have been using graph neural networks for traffic forecasting. David also mentioned an earlier study at Volkswagen called traffic flow optimization using quantum computers to essentially distribute the load across a road network graph.
Then, he emphasized the importance of being aware of societal risks and ethical concerns for every AI practitioner and data scientist:
“Whenever we use data or build AI, we should think about the implications and be mindful towards privacy, equitable ability, safety, transparency and security.”
Finally, David affirmed that AI R&D over the last 10 years has fueled the boom in applications and there are many untapped opportunities where we can leverage our AI superpowers for good.
Check out David von Dollen’s speech on our website