Vrije Universiteit Brussel

Tutorial about tactile sensing in robotics application

Hands-on Tutorial: 3D tactile sensor integration in robotic fingers for smart manipulation

Organisers: Milan Amighi and Bram Vanderborght (Brubotics - VUB), in collaboration with Melexis

Despite skin being our largest sensory organ, touch remains underused in robotics. Tactile sensing is vital for safe, dexterous manipulation, human-robot interaction, and wearables, yet most sensors measure only pressure. We live among materials with distinct rigidity, friction, weight and texture, differences humans intuitively feel thanks to their sense of touch. Robots can already move, see, and hear; the next step is to truly feel so they can grasp delicate, deformable objects while accounting for each object's properties.

This hands-on tutorial presents an affordable, compact, and robust 3D tactile sensor from Melexis: the Tactaxis® tactile 3D force sensor, which uses automotive-grade Hall-effect technology to measure both normal and shear forces. Participants will work in small teams to build a two-finger gripper, integrate the sensor, and deploy it on a real robotic task. Through guided experiments, you'll map force signals to interaction dynamics, detect slip, and enable adaptive control of the grasp depending on the object being handled. You'll leave with practical know-how, open-source resources, and a grounded understanding of why touch enables safer, more precise, and versatile robot behavior.

A first short lectures will introduce the Melexis 3D force sensor: its operating principles, underlying technology, advantages/limitations, and a brief state-of-the- art comparison, followed by a concise VUB talk on grasping theory (key challenges, design trade-offs, and best practices). We then move to application: contact modeling, slip detection, calibration, and impedance/admittance control. Attendees will build a simple two-finger gripper from scratch, integrate the sensor, and bring it up in ROS 2 (with a provided Docker setup), then implement an end-to- end manipulation pipeline with force-aware execution and test it on diverse objects. All materials (kits with 3D parts and fasteners), code, and a step-by-step guide will be provided in advance and on-site, with mentors available throughout. Attendees will leave with open-source CAD, code, and field-tested best practices for deploying touch-enabled manipulation. The session is designed to be accessible to a broad audience, from early-career to expert, so participants can grasp the principles and apply them to their own use cases, ultimately enabling more robust, safe, and dexterous robotic grasping in real-world environments. For the faster groups, it will be possible to adjust parameters in the software, such as those of the gripper and/or the sensor, so they can experiment with different settings and gain a deeper understanding of the concept.

Video of the Concept


Format

  • Presentation (60 min): Overview of 3D force sensing (principles, design trade- offs, relevance), then gripper-focused applications. Essentials of sensor integration/electronics and key grasping considerations (contact modeling, signal quality, latency, calibration). Run starter code to stream/visualize tactile data live. Ongoing Q&A on grasp stability, slip, and safety. Remote access via Microsoft Teams for people online will be available.
  • Coffee break (15 min)
  • Hands-on session (120 min):Teams of 3-4 build a two-finger gripper, integrate the sensor, and run experiments on objects with varied textures, weights, and fragility. Quick bill-of-materials check, then step-by-step assembly with notes on design choices and alternatives. Test, compare results, and discuss best practices for collecting/using physically grounded datasets. Q&A throughout. A step-by-step document is available for remote attendees; obtaining the full kit in advance may be challenging (sensors, 3D-printed parts, etc).

Contact information

Some references/papers about the sensors

  • T. Le Signor, N. Dupré and G. F. Close, "A Gradiometric Magnetic Force Sensor Immune to Stray Magnetic Fields for Robotic Hands and Grippers," in IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 3070-3076, April 2022, doi: 10.1109/LRA.2022.3146507
  • Le Signor, T.; Dupré, N.; Didden, J.; Lomakin, E.; Close, G. Mass-Manufacturable 3D Magnetic Force Sensor for Robotic Grasping and Slip Detection. Sensors 2023, 23, 3031. https://doi.org/10.3390/s23063031
Up

©2019 • Vrije Universiteit Brussel • Dept. MECH • Pleinlaan 2 • 1050 Elsene
• Tel.: +32-2-629.28.06 • Fax: +32-2-629.28.65 • webmaster