The Autonomous Driving Revolution and the Need for Accurate Simulations
Understanding the Complexity of Human Behavior in Autonomous Driving
The development and deployment of autonomous vehicles have the potential to revolutionize the transportation industry. However, one of the major challenges hindering progress in this field is the lack of accurate simulations of human behavior. Algorithms that can accurately reflect the behavior of road users, including pedestrians, motorists, and cyclists, are crucial for the safe and effective deployment of driverless vehicles.
A group of scientists led by Professor Gustav Markkula from the Institute of Transport Studies at the University of Leeds has identified the formidable complexity in developing software that can predict the way people behave and interact on roads. They emphasize that these accurate simulations are necessary to improve modeling and enhance the safety of autonomous vehicles.
The First Simulation of Human Behavior on Roads
To address this challenge, Professor Markkula and his research team have developed the first simulation that models how people behave on roads based on key cognitive theories. Instead of relying solely on statistical analysis or large data sets, the team focused on integrating computational psychological theories to create a more detailed and realistic model.
By integrating separate theories of psychology into a larger single psychological model, the researchers were able to describe behavior in more complex real-world tasks. During computer testing, the model accurately reproduced several well-known but previously ununderstood behaviors of pedestrians and drivers in common roadway scenarios. Additionally, the model predicted how real-life human subjects would behave when faced with interactive situations in a virtual reality simulator.
Unlocking the Potential of Autonomous Driving
Professor Markkula highlights that these findings shed light on the complex underlying cognitive mechanisms that influence everyday behavior of road users. Understanding the intricacies of human behavior is crucial for the successful development of autonomous vehicles.
The integration of separate psychological theories into a cohesive model for simulating human behavior on roads provides new perspectives on how to overcome the challenges in creating autonomous vehicles. Professor Markkula asserts that it is possible to develop models that accurately reflect the human dimension of behavior on the roads, which has been a long-standing request but rarely achieved.
The Importance of Accurate Simulations for Autonomous Vehicles
Economic Impact of Automated Vehicles
The development of automated vehicles not only has technological implications but also significant economic potential. The UK government recognizes this potential and has set ambitious goals to capitalize on the autonomous driving revolution. A vision statement from the government states that driverless vehicles have the potential to boost a £42 billion industry and create 38,000 new jobs. With such promising prospects, it is crucial to overcome the challenges that hinder the safe deployment of these vehicles.
The Role of Simulations in Testing and Development
Accurate models of human road user behavior are essential for testing and developing driverless vehicles and their control systems. Simulations that can demonstrate the vehicles’ safety in various scenarios involving human behaviors are crucial for gaining public trust and regulatory approval. While existing computer models have relied on statistical analysis, they often lack the level of detail necessary to accurately capture human behavior.
The research conducted by Professor Markkula’s team fills this gap by focusing specifically on the details of human behavior and key concepts of human psychology. By incorporating theories such as “theory of mind” and “behavioral game theory,” which explain how people form beliefs and consider the behavior of others, the researchers developed a model that better reflects the complexities of human interactions on the road.
Validating the Model Through Experiments
The researchers conducted experiments with human participants in the laboratory, including the use of a pedestrian simulator. These experiments revealed that the new model, based on psychological theory, could accurately predict driver-pedestrian interaction scenarios. This further validates the importance of integrating computational psychological theory into simulations to better understand and predict human behavior on the roads.
However, the researchers acknowledge that there is still much work to be done in developing models based on the psychology of road user behavior. The ultimate goal is to develop computer models that accurately reflect the human dimension of behavior on the roads, ensuring the safe and efficient integration of autonomous vehicles into our transportation systems.
Summary
The autonomous driving revolution is facing challenges due to the lack of accurate simulations of human behavior on roads. Algorithms that accurately reflect the behavior of road users are essential for the safe deployment of driverless vehicles. Recognizing this need, a team of researchers led by Professor Gustav Markkula from the University of Leeds has developed the first simulation of human behavior on roads based on key cognitive theories.
By integrating separate theories of psychology into a larger single psychological model, the researchers were able to accurately reproduce previously ununderstood behaviors of pedestrians and drivers. The model also predicted human behavior in interactive situations in a virtual reality simulator. These findings shed light on the complex cognitive mechanisms that influence everyday behavior on the roads and demonstrate the potential to develop accurate simulations for the safe deployment of autonomous vehicles.
—————————————————-
| Article | Link |
|---|---|
| UK Artful Impressions | Premiere Etsy Store |
| Sponsored Content | View |
| 90’s Rock Band Review | View |
| Ted Lasso’s MacBook Guide | View |
| Nature’s Secret to More Energy | View |
| Ancient Recipe for Weight Loss | View |
| MacBook Air i3 vs i5 | View |
| You Need a VPN in 2023 – Liberty Shield | View |
The autonomous driving revolution is hampered by a lack of accurate simulations of human behavior.
The scientists caution that algorithms that accurately reflect the behavior of road users, vital for the safe deployment of driverless vehicles, are not yet available.
They say there is “formidable complexity” in developing software that can predict the way people behave and interact on roads, whether they are pedestrians, motorists or cyclists.
To improve modelling, a research team led by Professor Gustav Markkula from the Institute of Transport Studies at the University of Leeds has developed the first simulation of how people behave on roads based on key cognitive theories.
Those separate theories were integrated into a larger single psychological model that would “describe behavior in more complex real-world tasks.”
During computer testing, the model accurately reproduced several well-known but previously ununderstood behaviors of pedestrians and drivers in common roadway scenarios. The model also predicted how real-life human subjects would behave when faced with interactive situations in a virtual reality simulator.
Professor Markkula said: “These findings suggest that the everyday behavior of road users is based on a number of complex underlying cognitive mechanisms, which may be part of the reason why it has been more difficult than expected to create autonomous vehicles. “.
“Our research shows that it is possible to integrate separate theories of psychology into combined theories for applications such as simulating the way people behave in traffic, which has been requested but rarely achieved.”
The researchers’ findings, which explain human interactions on the road by integrating computational psychological theory on a large scale, are published today (Tuesday, June 20) in the scientific journal PNAS Nexus.
Algorithms are needed to unlock the autonomous driving revolution
The development of automated vehicles could have a huge impact on the UK economy.
In a vision statement, the UK government has said driverless vehicles will boost a £42 billion industry and create 38,000 new jobs. The goal is to see the start of the safe deployment of driverless vehicles by 2025.
But writing in the scientific journal PNAS Nexus, the researchers argue that work towards driverless vehicles has been “hampered by a lack of models of how human road users interact.”
Accurate models are needed to run the necessary simulations both in the development and testing of driverless vehicles and their control systems, for example, to demonstrate that the vehicles remain safe when faced with a variety of human behaviors on the road. .
Until now, most computer models of road user behavior have been based on statistics, with predictions of how people would behave based on analysis of large data sets, but typically without analyzing those models at a level detailed behavior.
Instead, the research of Professor Markkula and his team has focused specifically on the details of human behavior and key concepts of human psychology.
Road user behaviors and theories
The researchers looked at several typical human behaviors that exist on the road, such as hesitation in unclear situations or implicit communication through vehicle or body movement to assert priority or encourage someone else to go first.
The model predicts how people will behave with reference to key cognitive theories. For example, one is “theory of mind,” in which people form beliefs about what another person does and how their own behavior can affect the decisions the other person makes. This is also related to “behavioral game theory”, which explains how people consider the combined effects of their own behavior and the behavior of others when deciding what to do.
Another theory incorporated into the model describes imperfect human perception, which requires people to take the time to assess and understand what is going on in their environment.
Tests with human participants in the laboratory, including the HIKER pedestrian simulator at the University of Leeds’ Virtuocity facility, revealed that the new model based on psychological theory could also make correct predictions about the driver-pedestrian interaction scenarios studied. in the experiments.
Professor Markkula, who holds the Chair of Applied Behavior Modeling at Leeds, added: “Our research has shown that by taking a number of existing but separate mathematical theories of human psychology and behaviour, and putting them together, we can model, in much more detail than was previously possible: how humans interact in road traffic, for example, as drivers or pedestrians, including phenomena such as hesitation and interpreting the intentions of others.”
In the paper, the researchers say that much work remains to be done in developing models based on the psychology of road user behaviour.
The overall goal, the researchers say, is to develop computer models that better reflect the human dimension of behavior on the roads.
The authors of the article, Explaining Human Interactions on the Road by Large-Scale Integration of Computational Psychological Theory, are Gustav Markkula, Yi-Shin Lin, Aravinda Srinivasan, Jac Billington, Matteo Leonetti, Amir Hossein Kalantari, Yue Yang, Yee Mun Lee, Ruth Madigan, and Natasha Mera. Matteo Leonetti is from Kings College London; the others are based at the University of Leeds.
https://www.sciencedaily.com/releases/2023/06/230620113736.htm
—————————————————-