London Tech Week 2026
Jun 25, 2026
The UK’s capital recently wrapped up London Tech Week 2026 at the Olympia Exhibition Centre, and the energy was electric. If you stopped by the newly debuted Sandbox area, a dedicated zone for live, cutting-edge tech demonstrations, you probably saw our team in action. Our showcase focused on our DEX-EE robotic hand and Shadow Tactile Sensors, supported by a real-time visual dashboard that made contact data visible and interpretable.
Headlines From London Tech Week 2026
For the last two years, the tech landscape has been heavily dominated by purely virtual AI, chatbots, large language models, and digital enterprise agents. But as the UK tech sector crosses a historic £1.2 trillion valuation, the frontier of innovation has firmly moved into Embodied AI.
The biggest news from London Tech Week 2026 was UK Prime Minister Keir Starmer announcing a massive £400 million investment plan to purchase specialist AI chips. The goal? To ensure the UK owns the foundational data infrastructure and processing capacity needed to power its own AI ambitions rather than relying entirely on overseas cloud monopolies.
DEX-EE in Action
Shadow Robot’s stand provided a rare opportunity for attendees to engage directly with advanced dexterous robotics hardware and tactile sensing technologies that connect digital systems to the physical world.
Shadow Robot’s DEX-EE robotic hand was at the centre of our demonstrations, highlighting its role as a robust and adaptable platform for dexterous manipulation research, teleoperation and learning-based robotics.
Visitors were able to interact with the system directly, observing how its three-finger architecture supports stable grasping, precise manipulation, and responsive control in real time. More importantly, the demonstrations sparked discussions around what makes robotics usable in real-world environments: not just form factor, but control fidelity, robustness, and the ability to generate meaningful training data under realistic conditions.
DEX-EE (available in both standard and Chiral configurations) is engineered specifically for high-cycle machine learning research and autonomous manipulation. Unlike traditional robotic hands designed for controlled environments, DEX-EE is built to withstand the realities of physical AI training, where systems are iteratively tested, refined, and sometimes pushed beyond expected operating conditions. This durability is complemented by active compliance at the joint level, allowing the hand to absorb unexpected forces while maintaining stable control performance under unstructured or high-torque behaviours produced by evolving robotic policies.
ShadowDash: From Sensors to Screen
One of the most engaging elements of the demonstration was our visual dashboard, designed to make robotic interaction data immediately accessible and showcase our Shadow Tactile Sensors on DEX-EE robotic hand.
DEX-EE includes multi-taxel, 3-degree-of-freedom tactile sensors on the distal, middle and proximal phalanges, providing rich additional information during grasping and manipulation. Each tactile module is user-replaceable and consists of 22 sensing cells on the proximal phalanx and 14 sensing cells on the middle phalanx. Each sensing cell integrates a magnet and a 3-axis hall effect sensor embedded within a compliant silicone surface. Together, they provide raw magnetic field measurements across normal (x) and shear (y, z) axes, along with temperature data, at 12-bit resolution. Sensor outputs update at 100 Hz per cell, enabling real-time tracking of contact forces, slip events, and surface interaction dynamics. This level of tactile resolution allows robots to move beyond vision-only perception and begin interpreting the physical quality of contact itself.
To make the captured tactile sensor data intuitive, we featured a dashboard built on the open-source RobotWhisperer framework. The visualisation dashboard extends far beyond translating tactile sensor outputs into visual feedback, as it is able to plot other sensor values such as IMU data and joint encoder positions as well. By aggregating this comprehensive hardware telemetry, the platform drives a virtual hand model that mirrors the physical robot’s movements and sensor readings in real time. This allowed visitors to see exactly how the robot experienced and responded to human touch.
Future of Tech
One of our favorite themes to unpack with visitors at our stand was a core principle we feel passionately about: the world doesn’t necessarily need flashy, walking humanoid bodies; it needs true, reliable dexterity.
While global tech conglomerates pour billions into making bipedal robots walk smoothly across flat warehouse floors, the most valuable, high-stakes tasks in industry require handling objects exactly like a human does.
True industrial scaling won’t happen because a robot can walk from Point A to Point B. It happens because a robot can delicately grip a glass vial, thread a complex assembly wire, or safely turn a pressure valve in a hazardous environment.
Missed us in the Sandbox? Explore our Dexterous Robotic Hands or get in touch with our team to find out how our teleoperation and manipulation systems can elevate your next project.