Reaching out: load cells for robotic datasets and other news
No sooner had we published our previous article on load cells and robotic grip than a new paper is published about load cells and hand kinematics data.
Autonomous robots rely on accurate datasets for their initial programming and set-up, and to tailor their grip to the required task. If the data only exists for gripping steel cans, and you want a robot to gently grip a crystal glass, for example, that can be a problem.
One of the other issues is the reaching phase, where the robotic hand travels towards the object. A team writing in Nature’s ‘Scientific Data’ journal noted that there were very few existing datasets for this ‘reach to grasp’ action. They wanted to create a data set that was “specifically tailored for autonomous grasping of a robotic hand.”
To do this the team needed to:
“Capture target object characteristics from radar and time-of-flight proximity sensors, as well as details of the reach-to-grasp action by looking at wrist and fingers kinematics, and at hand-object interaction main events.”
To collect data for all stages, researchers wore an instrumented glove fitted with multiple strain gauges and data acquisition electronics. They reached for and picked up a variety of standard shaped objects including spheres, cones and flat items. TAs the team explained:
“All objects were manipulated over one or two platforms each instrumented by a load cell. The load cells allow tracking of an action’s most salient events like touch, lift-off, replace and release.”
This approach delivering almost 6000 human-object interactions for the dataset, which the team hope will be useful for autonomous grasping robots in:
- health settings, assistive and/or rehabilitation robots
- industrial settings, including warehouse collaborative robots
- radar-based computer vision for recognition of objects
- sensorimotor control and manipulation
Round in circles
Load cells and laser sensors were at the centre of a row at the world-famous Le Mans 24 Hours race for 2023. According to an article in Autosport, no less than six P2 teams were given a 20minute stop-and-hold penalty for running illegal laser ride-height sensors during test runs.
According to the stewards’ bulletins, the use of sensors “gives a sporting advantage to the competitor”. However, Vector team principal Gary Holland said that the sensor data acquired was:
“Not a huge gain at all: we see it as a validation tool because everyone is running load cells on the push rods and damper potentiometers, plus of course everyone has a lot of historic data.”
Learning from the best
The International Rugby Experience in Limerick Ireland promises visitors ’six storeys of fully interactive, multifaceted digital engagement’ including individual and team-based skills trials. According to an article in Event Industry News,
“Visitors get to learn from legends including Jonny Wilkinson, Faf De Klerk, Keith Earls and Tadgh Furlong, via a ‘tech stack’ of touch screens, projectors, LiDAR sensors, laser spotlights and industrial load cells. “
Load cells are used across various sports for measuring, training and research, including neck strength testing fro rugby players , as discussed in our blog last year “Winning ways with load cells (Part 2)”
Looking for load cells for your innovative research or testing?
Contact us for expert help and advice, or browse our online shop.
