Models of the human body (crash test dummy, finite element models) are used in design and safety assessment of vehicles. However, these are limited to a few design points resulting in systems that are biased against sections of the population. A lack of experimental data and efficient tools historically restricted the scope of safety evaluations to a subset of road users. The current advances in digital resources, AI technology, and physics-aware computational methods open up new opportunities to design tools and workflows that can consider the diversity of road users in safety evaluations.In this study, we propose to develop a workflow to extract information from photographs to build individualized models and digital twins that can be used to assess injuries for wider sections of society (females, obese, elderly). In this project, we particularly focus on the development of a deep learning model to recognize anthropometry parameters, which would then be used to generate a digital twin representation of humans with appropriate anatomical landmarks and inertial properties. This workflow will enable users to combine existing empirical data from literature with person-specific information.

Chalmers participants:

• Shivesh Kumar (PI)

• Karinne Ramirez-Amaro (co-PI)

• Håkan Johansson (co-PI)

• Johan Davidsson (co-PI)

• Jobin John (co-PI)

Funding agency:

Chalmers Area of Advance Transport

Funding Chalmers participation during 2025-2026

Chalmers participants:

• Karinne Ramirez-Amaro (PI)

• Emmanuel Dean (co-PI)

Funding agency:

European Union (through the euROBIN network)

Cyclists pose incredible challenges for automated driving because their movements are hard to predict, and they may suffer severe injuries even from a low-speed collision with a motorized vehicle. To avoid collisions, cyclists and drivers negotiate intersections by predicting and showing each other’s intents. External interfaces may help automated vehicles showing their own intent but, unfortunately, will not help automation to predict human intent. Of course, a cyclist may keep away from an automated vehicle that shows its intent to pass first; however, this lack of interaction may not be safe nor acceptable. The alternative, automation always leaving the way to cyclists, is not acceptable (for the passengers in the automated vehicles) and is not necessarily safer either. The solution is teaching automation to predict human intent. This prediction can be done with interaction models, e.g. algorithms that describe how road-users influence each other intentions. Today, these models are still in their infancy, and their safety impact hard to prove because they are yet to be integrated into the safety assessment of the automated vehicle.

In this project, we will integrate interaction models into tools for the virtual safety assessment of automated vehicles, to prove the value of these models for transport safety. We will also use the simulations to derive the communication requirements for cooperative applications that may help the automated vehicle to have enough information about the environment to run the interaction models. To develop interaction models and their virtual safety assessment, we need real-world data. Therefore, this project will also support further development of a robot bicycle at E2 that will enable repeatable interaction experiments to collect the data that we need to develop and validate our interaction models and our tools for virtual safety assessment.

Chalmers participants:

• Marco Dozza (PI), Chalmers, Mechanics and Maritime Sciences (M2), Vehicle Safety

• Jonas Bärgman, Chalmers, Mechanics and Maritime Sciences (M2), Vehicle Safety

• Karinne Ramirez-Amaro, Chalmers, Electrical Engineering (E2), Systems and Control

• Jonas Sjöberg, Chalmers, Electrical Engineering (E2), Systems and Control

• Erik Ström, Chalmers, Electrical Engineering (E2), Communication, Antennas and Optical Networks

Funding agency:

Chalmers Area of Advance Transport

Funding Chalmers participation during 2021–2022

Human-robot collaboration is one of the most promising applications of autonomous robots. For this, robots need to interact with humans in a meaningful, flexible and adaptable manner, especially when new situations are faced. Currently, the new emerging technologies such as virtual reality and wearable devices allow capturing natural human movements of multiple users. Then, the next generation of learning methods should take advantage of and bootstrap the learning of new activities by adapting to the massive processing of information obtained from these enhanced sensors. The goal of this PhD project is to develop semantic-based learning algorithms able to cope with a large amount of data generated from multi-modal/multi-level sensors and react to dynamically changing environments in real-time to produce robots with enhanced autonomy levels and manipulation capabilities. The massive collection of data will also contain information about the different styles of human demonstrations which will help in developing a more human-centred control solution.

Chalmers participants:

• Karinne Ramirez-Amaro (PI)

• Yiannis Karayiannidis (co-PI)

• Jonas Sjöberg (co-PI)

• Maximilian Diehl (PhD student)

Funding agency:

Chalmers AI Research Centre (CHAIR)

Funding Chalmers participation during 2020-2025

Embodied Artificial Intelligence is a multidisciplinary area that requires the cooperation of different fields such as computer science, engineering, robotics and dynamical systems. This PhD thesis will develop a novel learning algorithm to allow high-level intelligence, such as problem-solving and reasoning to be applied to real-world physical systems, e.g. robots. Mainly, this work will be focused on investigating learning methods on the semantic aspects of intelligence to develop general purpose solutions for robotic applications. The goal of this PhD project is to develop compact and flexible model representations to allow robots the transference of their past experiences to current situations. These compact models should be human-readable to provide adequate feedback to the person interacting with the intelligent agent, e.g. a robot. This feedback should provide information to users, in an efficient form, about possible hazardous situations and potential errors, by predicting ahead-of-time the actions performed by either a human or a robot to infer their potential consequences. Thus, the communication between the users and the robots should be meaningful and bi-directional.

Chalmers participants:

• Karinne Ramirez-Amaro (PI)

• Jonas Sjöberg (co-PI)

• Wenhao Lu (PhD student)

Funding agency:

Chalmers AI Research Centre (CHAIR)

Funding Chalmers participation during 2020-2025