Understanding how behaviour emerges through mechano-ethology

01 · Instrumentation at the ant scale

A force platform to observe foot contacts

How can we measure the tiny forces exerted by each leg of a moving ant? This collaboration with Vincent FOURCASSIÉ and LAAS-CNRS combines biomechanics with optical sensor design. An array of deformable micropillars links camera-recorded displacements to forces in the plane of the substrate.

MANSARD et al. (2026) present a wide-field device for simultaneously measuring in-plane leg forces across several steps, with a reported resolution of 0.5 µN. The project favours an adaptable, robust and low-cost system. Co-authors: Vincent MANSARD, Elliot PERREZ, Dorian GAUTIER, Julien ROUL, Christian BERGAUD, Pierre MORETTO and Vincent FOURCASSIÉ. [1]

Observing foot contacts: Paul-Alain RAYNAL’s Master’s dissertation

This second-year Master’s dissertation documents foot contacts on micropillars and fluorescent tracking. [2] A technique to estimate the normal force component is under investigation.

Enlarge image

Read the publication on HAL ↗

02 · Rugby and team sports

Florent LOKTEFF — Quantifying impacts in rugby union

Thesis supervisor: Pierre MORETTO; co-supervisor: Saad DRISSI. Conducted through a CIFRE industrial PhD at Stade Toulousain, this thesis, defended in 2023, develops an impact-load index from wearable systems. It examines the strengths and limitations of accelerometry for characterising players’ mechanical loading. [3]

The paper by LOKTEFF et al. (2022) analyses two seasons at an elite club. Impact frequency and intensity vary by playing position: backs sustain fewer but more intense impacts than forwards. These findings support adapting load monitoring to positional demands. [4]

Enlarge imageEnlarge image

Thomas BÉCHARD — Linking collective indicators and individual contributions

Thesis jointly supervised by Pierre MORETTO and Cristian PASQUARETTA, through a CIFRE industrial PhD at Colomiers Rugby, and defended in 2025: Rugby union performance: linking collective indicators and individual impact. This research connects match analysis with team performance. [5]

BÉCHARD et al. (2025) examine 34 matches of a successful team, distinguishing line breaks from open play and set pieces. Dynamic line breaks are associated with a more favourable score difference both at home and away; the association for static line breaks depends more on match location. These results describe one team and one season, rather than establishing a universal rule. [6]

Arnaud ODET — Explaining strategies associated with winning

With Thomas BÉCHARD, Pierre MORETTO, Sébastien DÉJEAN and Cristian PASQUARETTA, ODET et al. (2025) propose a method combining machine learning and explainability. A model predicts outcomes from performance indicators; SHAP analysis then clarifies their role at three levels: overall trends, team strengths and weaknesses, and individual match diagnosis. The aim is to make predictions interpretable for analysts and coaches. [7]

A contribution to CAB 9 research

This work contributes to research on collective dynamics, which I have supported through doctoral supervision and collaborations leading to Cristian PASQUARETTA’s recruitment. My role in CAB 9 concerns (i) integrating impact detection and quantification into the analysis and prediction of collective movement in sport, and (ii) studying how feedback to players affects individual fitness and team organisation (Project TIRIS Sens-Perf 2026–2030, ongoing). SUSP call — Scaling Up Science Program ↗

Explore CAB 9 ↗

Comparing people of different sizes

Having two people walk or run at the same speed does not necessarily create mechanically equivalent tasks. Limb length, step frequency and energy exchange alter movement conditions. Dimensionless analysis seeks to separate size effects from differences in locomotor strategy.

This programme developed through doctoral research by Michaël BISIAUX on walking (2002), Nicolas DELATTRE on dynamic similarity in running (2008), and David VILLEGER on elastic energy return and locomotion (2014). Research byAntony COSTES on cycling (2016), co-supervised with Bruno WATIER, as was David VILLEGER’s thesis, extends the study of mechanical constraints and movement choices.

01 · Walking and running

FROUDE and STROUHAL: complementary information

The FROUDE number, usually written Fr = v²/(gL), compares inertial effects with gravity. It helps choose comparable speeds using a characteristic length, such as leg length. The STROUHAL number, St = fL/v, links movement frequency to speed and length, explicitly introducing step rhythm.

In a running study involving fifteen participants, DELATTRE, LAFORTUNE and MORETTO (2009) show that both numbers provide useful information, but neither alone achieves full dynamic similarity: STROUHAL particularly aligns temporal parameters, whereas FROUDE affects some mechanical parameters. This motivates combining them. [1]

Modela: linking rhythm and energy exchange

DELATTRE and MORETTO (2008) introduce a dimensionless number from exchanges between kinetic, gravitational potential and elastic energy in a spring–mass model. This formulation, which gave rise to Modela, links the energy balance to a combination of FROUDE and STROUHAL. Initially a modelling result, it was subsequently tested experimentally. [2]

Research by VILLEGER, COSTES, WATIER and MORETTO compares three conditions in nineteen participants: identical speed, FROUDE-adjusted speed, and jointly adjusted speed and frequency. Modela-r for running (2014) and Modela-w for walking (2015) improve similarity and reduce variability in dimensionless mechanical parameters. In the tested protocols, the third condition met similarity criteria for 52 of 54 running parameters and 30 of 36 walking parameters, versus 32 and 22 respectively with FROUDE alone. [3, 4]

Building on Modela’s dimensionless approach, MORETTO et al. (2016) clarify energy exchange in a multi-segment system by relating changes in the spring–mass model’s elastic energy to changes in internal kinetic energy minus the work of internal forces. This highlights how coordination of joint work contributes to an emergent overall stiffness. [8]

12

The aim is to compare techniques, equipment or populations under better-matched mechanical conditions. These results apply to the tested protocols and do not, by themselves, constitute diagnostic validation.

02 · Cycling

Why stand up on the pedals?

The thesis byAntony COSTES examines optimisation strategies and the seated-to-standing transition on a cycle ergometer, including fatigue effects. The selected papers analyse forces on the bicycle and whole-body coordination as power increases.

COSTES et al. (2015) instrument the saddle, pedals and handlebars and reconstruct body movement. In 25 non-cyclists, increasing pedal forces coincide with a reduction in minimum vertical saddle force. Near the spontaneous transition, handlebar and pedal actions counter upward trunk motion. The findings suggest that saddle unloading contributes to standing up. [5]

123

Upper-limb analysis also shows that wrist, elbow and shoulder loading increases with power and is greater when standing: pedalling requires a whole-body perspective. [6] Another study links spontaneous transition to minimising cost functions based on joint moments and muscle activity: sitting is more favourable below transition power, standing above it. [7]

Agreement between electromyography and biomechanics. Muscle-activity cost functions, including muscle-volume weighting, and joint-moment cost functions reveal the same trend: sitting is more favourable at low power, whereas standing gains the advantage at high power. This agreement suggests that the transition reorganises mechanical and muscular loading as power increases. [7]

The connection with walking and running is the search for mechanical principles explaining how a strategy emerges. These cycling papers do not validate Modela for pedalling.

A foundation for CAB 10

This work provides two complementary tools: comparing systems across scales and identifying constraints that shape movement. CAB 10 extends this approach to ants, robotic legs and exoskeletons, checking physical assumptions and material properties at every transfer.

Another branch concerns measuring foot contacts and providing feedback: how can a biomechanical measurement become information that helps a person adapt their movement?

Making foot loading perceptible

Baropodometry measures plantar pressure distribution. Biofeedback adds a feedback loop: visual or auditory signals inform people about foot loading and help them voluntarily adjust their gait. The aim is to unload a target area while monitoring pressure redistribution elsewhere under the foot.

Doctoral research by Virginie FEMERY (2003), followed byAurélien DESCATOIRE (2009), at Lille 2, helped develop and evaluate this approach. In Toulouse, research by Laëtitia CLAVERIE (2017), through a CIFRE industrial PhD with Médicapteurs, co-supervised with Anne ILLE, focused on a wearable system for assessing locomotion from plantar pressure.

01 · FEMERY and DESCATOIRE

BIRD: feedback to change foot loading

The study by FEMERY et al. (2008) investigates voluntary partial unloading of the first metatarsal head using visual and auditory signals. Five of eight healthy participants significantly changed their foot loading as requested. However, excessive unloading highlights the need for learning and precise feedback. [1]

BIRD — Baropodometric Information Return Device combines plantar sensors and pressure-threshold feedback. DESCATOIRE et al. describe the device and demonstrate, in feasibility studies, changes in gait strategy to unload a target area. A clinical aim is to provide substitute information when protective foot sensation is impaired. [2, 3]

Enlarge image

These studies establish the feasibility of plantar pressure feedback. They do not, by themselves, demonstrate reduced ulcer incidence or long-term clinical effectiveness.

02 · CLAVERIE and Médicapteurs

Well-positioned sensors for wearable measurement

A wearable system must stay light and simple while accurately describing foot loading. Research by CLAVERIE, ILLE and MORETTO asks where a limited number of sensors should be placed under the foot. The 2015 contribution presents the method; the 2016 paper compares two sensor layouts in the W-inshoe® system against a force platform. [4, 5]

In ten healthy volunteers, the adjusted layout correlates well with the force platform for centre-of-pressure displacement and vertical force. Measurement quality therefore depends on sensor location, not just number. This validation opens possibilities for monitoring outside the laboratory and real-time feedback. [5]

Enlarge image

Laëtitia CLAVERIE’s thesis forms part of this collaboration with Médicapteurs. The activity report also describes Olivier GUTIERREZ’s complementary work on auditory gait feedback, linking instrumentation, perception and movement adaptation.

03 · Technology transfer and rehabilitation

Serious GaRS: foot loading as a game interface

Serious Games for Rehabilitation in Seniors — Serious GaRS extends the collaboration with Médicapteurs through regional and European ERDF funding. Developed in response to theEASYNOV 2016 — Silver Economy call, the project was selected for funding reported as 350 k€ in the activity report.

The report describes pressure platforms linked to a gaming environment: participants’ movements control an avatar interacting with a virtual group. Research on collective behaviour, particularly by Guy THERAULAZ and Mathieu LIHOREAU, informs this combination of biomechanics, physical activity and gaming.

The aim is to offer engaging exercises for older adults’ rehabilitation and fall prevention. The report presents prototype development and planned assessments; these objectives are distinguished here from published instrumental validation of the measurement systems.

Explore EASYNOV at Médicapteurs ↗

Enlarge photoEnlarge photoEnlarge photo