Épisodes

  • Carmen Amo Alonso: Part 2 (AI Architectures and the Dynamics of Intelligence)

    Carmen Amo Alonso: Part 2 (AI Architectures and the Dynamics of Intelligence)
    Jun 7 2026
    In the second part of my interview with the incredibly talented Carmen Amo Alonso we discuss her fascinating work at the intersection of AI and control, examining topics such as:Using tools from control theory to analyse and design AI systemsControlling large language models to guarantee a particular sentiment for their outputUsing vision-language-action models to control robots using languageand a whole lot more! You can find the transcript for this episode at: https://docs.google.com/document/d/1ph1yaQuV_ybRM8h9rU5Uvmz9eB5Ar4HVhcBnc7b7zJ8/edit?usp=sharingFollow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/ ReferencesThe architecture of language: https://www.youtube.com/watch?v=ApC0AWWTajUDiffusion models: https://the-principles-of-diffusion-models.github.io/Skip connection paper: https://arxiv.org/abs/2410.10609 Review of state-space models in AI: https://arxiv.org/abs/2403.16899 3 Blue 1 Brown video on transformers: https://www.youtube.com/watch?v=wjZofJX0v4MDynamical systems framework paper: https://arxiv.org/abs/2405.15731 Designing sequence models: https://arxiv.org/abs/2510.09389v1Controlling LLM output: https://arxiv.org/abs/2405.15454v3Controllability of LLMs: https://arxiv.org/abs/2601.05637v1 Vision-language-action model paper: https://arxiv.org/abs/2603.05487v1NARRATE: https://arxiv.org/abs/2403.10762DEMONSTRATE: https://arxiv.org/abs/2507.12855v1Domitilla Del Vecchio: https://meche.mit.edu/people/faculty/ddv@MIT.EDUBiomolecular feedback systems book: https://www.fbswiki.org/wiki/index.php/Biomolecular_Feedback_SystemsMathias Foo episode: https://open.spotify.com/episode/1G3vhRoeenTWRrEwauT0neinControl episodes: Mario di Bernardo: https://www.incontrolpodcast.com/1632769/episodes/19167890-ep44-mario-di-bernardo-from-circuits-to-cells-and-swarms-control-meets-complexityMustafa Khammash: https://www.incontrolpodcast.com/1632769/episodes/12648716-ep11-mustafa-khammash-cybergeneticsJohn Doyle: https://ieeecss.org/contact/john-doyleInControl episode: https://www.incontrolpodcast.com/1632769/episodes/12841020-ep12-john-doyle-part-i-a-pioneer-s-guide-to-robust-control-the-past-present-and-futureJerome Sieber: https://jerome.sieber.io/
    Afficher plus Afficher moins
    52 min
  • Carmen Amo Alonso: Part 1 (Distributed and Localized Model Predictive Control)

    Carmen Amo Alonso: Part 1 (Distributed and Localized Model Predictive Control)
    Jun 2 2026

    In this episode I talk to the very wonderful Carmen Amo Alonso about her (award-winning) PhD research on distributed and localised MPC. We also cover super interesting and useful concepts such as system-level synthesis and ADMM, as well as her internship working at Tesla!


    You can access the transcript at: https://docs.google.com/document/d/1SErvHtKoE18QzelHDUh9im6AeG3zKPWZLRoNGx5BUyg/edit?usp=sharing


    Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/

    Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/

    References

    Carmen: https://camoalon.github.io/

    JPL: https://www.jpl.nasa.gov/

    John Doyle: https://ieeecss.org/contact/john-doyle

    Feedback in biological control: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9867859

    https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9867794

    Thesis: https://thesis.caltech.edu/15262/3/Carmen_Amo_Alonso_Caltech_PhD_Thesis.pdf

    Distributed localised MPC paper: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9939038

    Professor Melanie Zeilinger: https://nccr-automation.ch/about/people/melanie-zeilinger-0

    SLS for robust nonlinear optimal control: https://arxiv.org/pdf/2301.04943

    Professor Nikolai Matni: https://nikolaimatni.github.io/

    System-level synthesis tutorial: https://arxiv.org/abs/1904.01634

    YouTube videos on DLMPC: https://www.youtube.com/watch?v=mF2Wbriy2Zg

    https://www.youtube.com/watch?v=IhtuGZp8QM0&t=302s

    ADMM survey paper: https://arxiv.org/abs/2208.03700

    Professor James Anderson: https://www.columbia.edu/~ja3451/

    Data-driven DLMCP: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9772975

    Willems fundamental lemma: https://eprints.soton.ac.uk/262195/1/PersistencyExcitation.pdf

    Paolo Rapisarda episode: https://open.spotify.com/episode/076dEw28abILPAbT5qA6aw

    Implementing DLMP on GPUs: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9831839

    Clubes de Ciencia: Mexico: https://clubesdeciencia.mx/

    Peru: https://clubesdecienciaperu.org/

    Terry Sejnowski: https://biology.ucsd.edu/research/faculty/tsejnowski
    Afficher plus Afficher moins
    30 min
  • Journal Clubs (Creating reading groups that people want to be part of)

    Journal Clubs (Creating reading groups that people want to be part of)
    May 1 2026

    In this episode I talk to three lovely researchers all about journal clubs (AKA reading groups) – what they are, why you might want to be part of one, and how to start your own! I also give some insights from my own experiences running a (control-related) journal club, as well as a summary of three of my favourite papers we’ve read.

    You can find the transcript at: https://docs.google.com/document/d/19i4egf_2KEWqkp94ZW3QjRTH8PeAZz6yEix8zoSfx4k/edit?usp=sharing

    Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/

    Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/

    References

    Al Edwards: https://www.southampton.ac.uk/people/62mm7y/doctor-al-edwards

    Antonia Marcu: https://www.southampton.ac.uk/people/5xk5lz/miss-antonia-marcu

    Convexification for soft landing optimal control paper: http://www.larsblackmore.com/iee_tcst13.pdf

    Improved soft robotic parameterisation paper: https://ieeexplore.ieee.org/document/8961972

    KKL observer synthesis paper: https://arxiv.org/abs/2501.11655

    Afficher plus Afficher moins
    34 min
  • Brian Douglas (Control Engineering in Industry)

    Brian Douglas (Control Engineering in Industry)
    Apr 1 2026

    In this episode I got the opportunity to sit down with the fantastic Brian Douglas and talk about what it's like to be a control engineer in industry, common misconceptions he had, and advice for current students. We also discussed his YouTube channel, as well as a few more technical aspects of projects he's been involved with over the years!


    You can find the transcript at: https://docs.google.com/document/d/1gAyIAgd3RzVSSDmJ-As0gzJVekJG3IqU2L-1zJhuKYs/edit?usp=sharing

    Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/

    Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/

    References

    Brian Douglas: https://www.linkedin.com/in/brian-douglas-505b7175/

    Engineering Media (Brian’s company): https://engineeringmedia.com/

    Original YouTube channel: https://www.youtube.com/channel/UCq0imsn84ShAe9PBOFnoIrg

    MATLAB YouTube: https://www.youtube.com/@MATLAB

    Co-design podcast episode: https://open.spotify.com/episode/3upRUgZBy79tqihMy9BDsJ


    Veritasium: https://www.youtube.com/@veritasium

    Smarter every day: https://www.youtube.com/channel/UC6107grRI4m0o2-emgoDnAA

    Mark Rober: https://www.youtube.com/@MarkRober

    Formal verification video: https://www.youtube.com/watch?v=kINfn-ZYiLY

    InControl episode: https://www.incontrolpodcast.com/1632769/episodes/15223020-ep24-brian-douglas-boeing-control-videos-resourcium-map-of-control-theory-cartoons-mathworks
    Afficher plus Afficher moins
    59 min
  • Gioele Zardini (Engineering Trade-offs and Co-design)

    Gioele Zardini (Engineering Trade-offs and Co-design)
    Mar 1 2026
    In this episode I got to speak with the fantastic Gioele Zardini about his work on the co-design framework, a fascinating framework used to optimise large interconnected systems. We cover the theoretical background of this, intuition behind it, and a ton of different applications where it’s useful, as well as discussing exciting open questions in the field! Find the transcript at: https://docs.google.com/document/d/1XPZr9jP4Lphskkw-SiM8h-x1kp7LROqZ243S6zA7EsY/edit?usp=sharingFollow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/ ReferencesGioele Zardini: https://zardini.mit.edu/PhD thesis: https://www.research-collection.ethz.ch/entities/publication/d7c08dd5-bf96-4c1f-a744-5e751f0f44a5PhDs in control interview: https://ieeexplore.ieee.org/document/10555200Emilio Frazzoli: https://idsc.ethz.ch/research-frazzoli/people/person-detail.frazzoli.htmlAndrea Cenci: https://nccr-automation.ch/about/people/andrea-censiMarco Pavone: https://profiles.stanford.edu/marco-pavoneNuTonomy: https://en.wikipedia.org/wiki/NuTonomy(current site: https://motional.com/)Brad Nelson: https://msrl.ethz.ch/the-lab/team/Brad_Nelson.htmlDavid Sacramuza: https://rpg.ifi.uzh.ch/people_scaramuzza.htmlYujun Huang: https://zardini.mit.edu/people/yujun-huang/Uncertainty in co-design paper: https://arxiv.org/pdf/2504.02766 Background refs on category theory:David Spivak: https://dspivak.net/Brendan Fong: http://brendanfong.com/ Gioele's work in progress book: https://storage.zuper.ai/sync/ACT4E/ACT4E/alphubel-prod/build/last/build-public/ACT4E-public.pdfYouTube talk: https://www.youtube.com/watch?v=BmAz6fEPa2oGioele's course: https://zardini.mit.edu/act4ed/MIT applied category theory course: https://ocw.mit.edu/courses/18-s097-applied-category-theory-january-iap-2019/Seven sketches in compositionality: http://brendanfong.com/fong_spivak_an_invitation.pdf LQG control co-design paper: https://ieeexplore.ieee.org/abstract/document/9654960LQG background: https://www.youtube.com/watch?v=H4_hFazBGxUhttps://www.youtube.com/watch?v=E_RDCFOlJx4Aaron Ames: http://ames.caltech.edu/Nikolai Matni: https://nikolaimatni.github.io/John Doyle: https://www.eas.caltech.edu/people/doyleCDC episode: https://open.spotify.com/episode/3ZUcm3gVHKILDEyDjQtXvx Braess’s paradox: https://resources.mpi-inf.mpg.de/departments/d1/teaching/ws12/ct/Braess-paradox.pdf Co-design of mobility-on-demand systems: https://arxiv.org/pdf/2003.04739 Formula One co-design papers: https://www.research-collection.ethz.ch/entities/publication/bd65f78e-f19a-4469-a379-d4b219893a33https://CoffeeControlLinks.short.gy/JQmudBCo-design of soft robotics: https://arxiv.org/pdf/2505.03761 Cosimo Della Santina inControl episode: https://CoffeeControlLinks.short.gy/LPNORz Games with partially-ordered objectives: https://arxiv.org/pdf/2111.07099 Mingjia He: https://mingjia-he.github.io/Paper on hierarchical games: https://arxiv.org/pdf/2508.12059 Game theory episode: https://open.spotify.com/episode/2xrwua7U2HfSW3sxY6BnRUJeff Shamma inControl episode: https://www.incontrolpodcast.com/1632769/episodes/18482182-ep40-jeff-shamma-gain-scheduling-nonlinear-control-learning-dissipativity-in-games-jiu-jitsuVincent Abbott: https://zardini.mit.edu/people/vincent-abbott/Alberto Sangiovanni-Vincentelli: https://www2.eecs.berkeley.edu/Faculty/Homepages/sangiovanni-vicentelli.htmlInControl episodes: https://www.incontrolpodcast.com/1632769/episodes/14692449-ep21-alberto-sangiovanni-vincentelli-part-i-classics-eda-logic-synthesis-spice-intel-x386-cadence-synopsis-platform-based-designhttps://www.incontrolpodcast.com/1632769/episodes/14882529-ep22-alberto-sangiovanni-vincentelli-part-ii-cyberphysical-and-hybrid-systems-contracts-chiplets-models-of-computation-time-innovative-ecosystems-biobricks-neuromorphicsJohn Doyle: https://www.eas.caltech.edu/people/doyle
    Afficher plus Afficher moins
    1 h et 10 min
  • Barbara Franci (Game Theory and Resource Allocation)

    Barbara Franci (Game Theory and Resource Allocation)
    Feb 1 2026
    In this episode I have a lovely conversation with Dr Barbara Franci about her research into game theory, exploring both theoretical concepts (such as variational inequalities, forwards-backwards algorithms, and stochastic Nash equilibrium-seeking problems) and practical applications, including resource allocation, energy management, and generative adversarial networks! You can find the transcript at: https://docs.google.com/document/d/1h9AbCPzzlt7ggNOGx3bK3pwSaCUo__LtYzvdJ-5wqYw/edit?usp=sharing Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/ ReferencesBarbara Franci: https://www.maastrichtuniversity.nl/b-franciPhD thesis: https://tesidottorato.depositolegale.it/bitstream/20.500.14242/198173/1/FRANCI%20Tesi2.pdfMore background on evolutionary games: https://users.ssc.wisc.edu/~whs/teaching/805/handbook.pdfJeff Shamma inConctrol episode: https://www.incontrolpodcast.com/1632769/episodes/18482182-ep40-jeff-shamma-gain-scheduling-nonlinear-control-learning-dissipativity-in-games-jiu-jitsu Wisdom of a crowd paper: https://epubs.siam.org/doi/abs/10.1137/18M1232267Francesco Bullo: https://fbullo.github.io/Original Galton paper: https://www.nature.com/articles/075450a0 Video on the Lagrangian: https://www.youtube.com/watch?v=GR4ff0dTLTwUseful lecture notes on these concepts: https://www.mit.edu/~gfarina/notes/ Distributed forwards-backwards paper: https://ieeexplore.ieee.org/abstract/document/9309063Extragradient paper: https://ieeexplore.ieee.org/abstract/document/11184508Relaxed inertial forwards-backwards paper: https://ieeexplore.ieee.org/abstract/document/9682905 Professor Lacra Pavel: https://www.ece.utoronto.ca/people/pavel-l/Deterministic decoupling paper: https://www.researchgate.net/publication/330888828_An_operator_splitting_approach_for_distributed_generalized_Nash_equilibria_computationPseudo-monotone functions: https://www.researchgate.net/publication/225986571_Pseudomonotone_Operators_A_Survey_of_the_Theory_and_Its_ApplicationsInertial gradient descent: https://www.youtube.com/watch?v=iudXf5n_3ro&t=60s GANs paper: https://ieeexplore.ieee.org/abstract/document/9523601Original Goodfellow paper: https://arxiv.org/abs/1406.2661 Partial information paper: https://www.sciencedirect.com/science/article/pii/S0005109821006300Incorporating a bias: https://ieeexplore.ieee.org/abstract/document/10886508Nash games with misleading information: https://ieeexplore.ieee.org/abstract/document/11008595Evolutionary dynamics in distributed control: https://ieeexplore.ieee.org/document/7823106
    Afficher plus Afficher moins
    55 min
  • Meet the Faculty Candidates (Fascinating research from the CDC poster session)

    Meet the Faculty Candidates (Fascinating research from the CDC poster session)
    Jan 9 2026
    In this bonus episode I explore the work of ten fascinating researchers that I met at the Meet the Faculty Candidates poster session at CDC last month. We cover a wide array of different areas, including data-driven methods, fault-tolerant control, biological systems, PDEs, distributed estimation and a whole lot more! Correction: At ~5:20 I mention that Anju was able to characterize the strategies of the two communicating entities at the Nash equilibrium - In actual fact this is a Stackelberg equilibrium (Stackelberg games involve players acting sequentially, taking either a leader or follower role)You can view the transcript for this episode at: https://docs.google.com/document/d/15Tf4ORHPI5Thbs-nA9DQTSu87TKbYni9JJN2UXqn5kU/edit?usp=sharingFollow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/ Links and referencesAmy Strong: https://bridgeman.pratt.duke.edu/people/amy-strongPaper on data-driven invariance: https://arxiv.org/pdf/2511.19231 Anju Anand: https://sites.google.com/binghamton.edu/anjuanandStrategic quantisation: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10960399Incorporating privacy into the objective: https://arxiv.org/pdf/2506.06321 Brian Block: https://mae.osu.edu/people/block.168Control of PDE traffic flow: https://arxiv.org/pdf/2512.04823 Declan Jagt: https://search.asu.edu/profile/3729730Stability analysis of coupled multivariate PDEs: https://arxiv.org/pdf/2508.14840 Scalar quadratic PDEs: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10384073PIETOOLS software: https://control.asu.edu/pietools/pietoolsHamza El-Kebir: https://www.linkedin.com/in/hamza-el-kebir/Control authority degradation paper: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10354437Sensor and actuator degradation: https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=10521243Electrosurgery paper: https://royalsocietypublishing.org/rsif/article/21/210/20230420/90578/Heat-conduction-in-live-tissue-duringElectrostatic signed distance functions: https://arxiv.org/pdf/2508.12554 Kirill Sechkar: https://engbio.ox.ac.uk/people/kirill-sechkarControl of genetic circuits paper: https://www.biorxiv.org/content/10.1101/2025.04.01.646303v2.full.pdf Nick Marios Kokolakis: https://www.linkedin.com/in/nick-marios-t-kokolakis-44717b164/Pre-defined time RL paper: https://www.sciencedirect.com/science/article/abs/pii/S0005109825003152 Parth Paritosh: https://pptx.github.io/pparitosh/Estimating a subset of variables: https://arxiv.org/pdf/2312.01227Privacy preserving paper: https://ieeexplore.ieee.org/abstract/document/11073227Radoslaw Patelski: https://www.linkedin.com/in/radoslaw-patelski/?originalSubdomain=plObservers for hovercraft: https://ieeexplore.ieee.org/abstract/document/9874346ADRC with parameter identification: https://ieeexplore.ieee.org/abstract/document/10345678ADRC with adaptive input gain: https://www.sciencedirect.com/science/article/pii/S0019057825001089 Carmen Amo Alonso: https://camoalon.github.io/Research overview: https://camoalon.github.io/research/Comparing mixer blocks: https://arxiv.org/pdf/2405.15731 Intro to state space models in deep learning: https://arxiv.org/pdf/2403.16899 Combining generative AI with control in robotics: https://arxiv.org/pdf/2403.10762
    Afficher plus Afficher moins
    26 min
  • CDC special (Conference of Decision and Control 2025)

    CDC special (Conference of Decision and Control 2025)
    Jan 1 2026
    In this episode I bring my microphones with me to the conference of decision and control (CDC) where I talk to a load of interesting researchers about their recent work. Featuring discussion of a wide range of different topics, including reinforcement learning, control architecture, game theory, AI-human interaction, event-triggered control and a whole lot more! You can view the episode transcript at: https://docs.google.com/document/d/1h6PdTd9xLTcJOGPx6TIkXOKPz_NGoS9_sA8CUvoZlx8/edit?usp=sharing Follow the podcast on LinkedIn: https://www.linkedin.com/company/coffee-and-control-podcast/Or connect with me directly: https://www.linkedin.com/in/lucy-hodgins-733a30175/ CDC web-page: https://cdc2025.ieeecss.org/Female influencers in control: https://www.control.lth.se/external-engagement/female-influencers-in-automatic-control/Survey: https://docs.google.com/forms/d/e/1FAIpQLSe7VtNHI6bG8UBEoSGiAuVXNsmXSHlw8xG0LuMon2DVd0Pv0A/viewform inControl episode: https://www.incontrolpodcast.com/1632769/episodes/18361436-ep39-female-influencers-in-control Alessandro Abate: https://www.cs.ox.ac.uk/people/alessandro.abate/Maegan Tucker: https://maegantucker.com/Joe Moeller: https://joe-moeller.com/Category theory blog: https://joe-moeller.com/hi/Rodolphe Sepulchre: https://www.eng.cam.ac.uk/profiles/rs771inControl episode: https://CoffeeControlLinks.short.gy/9HLW37Spiking control of a snake: https://arxiv.org/pdf/2509.02968 Neuromorphic architecture: https://arxiv.org/pdf/2511.11924 Raj Velichete: https://ece.illinois.edu/about/directory/grad-students/rkv4Vijeth Hebber: https://www.vijethhebbar.com/John Doyle: https://www.eas.caltech.edu/people/doyleInControl episodes: https://CoffeeControlLinks.short.gy/yx0IF5https://CoffeeControlLinks.short.gy/qtk0LeDimos Dimarogonas: https://people.kth.se/~dimos/Plenary on formal verification for multi-agent systems: https://www.youtube.com/watch?v=5C3GSgVmxdo&t=1327sAttacks on controlling multi-agent systems: https://ieeexplore-ieee-org.soton.idm.oclc.org/stamp/stamp.jsp?tp=&arnumber=11036540ESC inspired by insect flapping: https://ieeexplore-ieee-org.soton.idm.oclc.org/stamp/stamp.jsp?tp=&arnumber=11006034My CDC paper: https://pure.soton.ac.uk/admin/files/315205732/CDC2025_2_.pdfJorge Poveda: https://poveda.ucsd.edu/Victoria Grushkovskaya: https://grushkovska.science/ Higher-order Lie bracket extremum-seeking: https://arxiv.org/pdf/2310.07092 https://arxiv.org/pdf/2504.01136 Azu Ozdaglar: https://asu.mit.edu/Full video: https://www.youtube.com/watch?v=2DR37zdEKPwMatt Turner: https://www.southampton.ac.uk/people/5y5nkg/professor-matthew-turnerMatt Turner episode: https://open.spotify.com/episode/5KLLOvbCjOkL0v8iXo84vtKemin Zhou: https://www.ece.lsu.edu/kemin/Multiobjective complementary control: https://ieeexplore-ieee-org.soton.idm.oclc.org/document/10767397 Yujun Huang: https://zardini.mit.edu/people/yujun-huang/Tyler Hanks: https://gataslab.org/members/tylerhanksNextCom: https://sites.google.com/view/css-nextcomNextCom podcast episode: https://open.spotify.com/episode/7exZHRwVazI8AaHtiSMmE8Csaba Szepesvari: https://sites.ualberta.ca/~szepesva/Friday plenary talk: https://www.youtube.com/watch?v=-W2IrHyuSjUFengyu Yue: https://www.linkedin.com/in/fengyu-yue-8849a6314/?originalSubdomain=ilPio Ong: https://pioong.github.io/Jacqueline Scherpen: https://www.rug.nl/staff/j.m.a.scherpen/?lang=enBode lecture: https://www.youtube.com/watch?v=-W2IrHyuSjU (Skip to 4:52:46)Alberto Padoan episode on model order reduction: https://open.spotify.com/episode/1kwlnlQqtID1naRcyLL12OMeggie Maschner: https://www.linkedin.com/in/meggie-marschner-445634236/?originalSubdomain=deBrooks Butler: https://sites.google.com/view/brooks-butlerMauro Salazar: https://www.tue.nl/en/research/researchers/mauro-salazarDesign of smart cities: https://www.youtube.com/watch?v=-W2IrHyuSjU (Skip to 8:32:49)
    Afficher plus Afficher moins
    51 min