M1/M2 : Modeling & analyzing human biomechanical variability to inform design

Auctus team – Inria Bordeaux
Contacts :
Pierre Puchaud, pierre.puchaud@inria.fr
Margot Vulliez, margot.vulliez@inria.fr

Research environment
This research internship will take place within the Auctus team at the Inria center of the University of
Bordeaux (Talence). The Auctus team aims at meeting challenges of collaborative robotics for humans
at work. The team's research is divided into three scientific axes: analysis and modeling of human
behavior (biomechanical and cognitive); human-robot interaction and coupling; design and control of
cobotic systems. This internship falls at the intersection of the first and third axis, by participating in
the definition of human-centered specifications of assistive device, specifically exoskeleton. The project
will benefit from the hardware and software resources of the lab experimental platform. The student
will be supervised throughout the internship by the two supervisors.

Research activities
To build some insights on human-centered design requirements for an adaptive upper-limb
exoskeleton, the internship will be structured in two parallel research axes combining biomechanical
modeling and mechanisms analysis:

- Existing statistical models of human variability.
The intern will read scientific articles and look for existing and open-sourced statistical human models,
such as SMPL-X [Pavlakos19], that propose to modify the morphology and kinematic chain through
“principal components”. The intern will present his/her readings and test and install the associated
software.
The intern will then compare the effect of principal components and axes of variation on the statistical
human model among existing sotfware.
Based on this analysis, the intern will generate a set of human envelops (that represents a population)
and will convert it into biomechanical kinematic chains [Keller23] in a multibody simulation engine. We
will extract upper-limb parts into a Pinocchio-modeling [Carpentier19] compatible format in order to
study the kinematic variability to inform exoskeleton design.
- Requirements for adaptation solutions of exoskeleton design.
The intern will do a state of the art of exoskeleton design to survey existing technical solutions to adapt
the device to user variability. Common solutions, as survey in [Näf18, Chen24], can be to add redundant
and passive degrees of freedom to the kinematic chain [Manzano25], to design self-alignment
mechanisms [Cempini12, Schorsch14], or to integrate compliance through flexible/soft mechanisms at
the anchor point or within the mechanical structure [Lee16].
The existing adaptation solutions will be compared in terms of possible adjustment levels: degrees of
freedom, direction and range of adjustment.
The intern will then extract from the human variability models recommendations to choose the number
and axis direction of the adaptation degrees of freedom needed for an elbow exoskeleton. Extra
settings (parameters) could also be considered for more adjustments.
- Design optimization (optional)
The intern will optimize a set of parameters of the adaptation mechanisms of a given elbow exoskeleton
to fit the kinematics of a user and a selected whole population. Existing design optimization
methodologies [Sposito20, Zhou17] can be used to guide the optimization process.
The optimized exoskeleton will be compared to the initial one with respect to mobility and comfort
metrics.

Skills
The candidate should have solid skills in kinematic and dynamic modeling, data analysis, and
programming (C++ or Python). Additional experience in biomechanics, statistics, machine-learning, or
mechatronic design or experience with git would be appreciated.

Instruction to apply
To apply at this offer, please email a detailed CV, a motivation letter, and your Master courses/grades,
to the two supervisors. If you have a portfolio of projects (software or design) don’t hesitate to transmit
them too.