A team of researchers at the University of Melbourne wants to make the interaction between man and machine even more interactive and fluid. To achieve this, they have developed a “predictive touch response” mechanism that greatly optimizes touch sensitivity in machine control.
The development of the predictive touch response mechanism is the next step in machine control. According to current predictions, especially the Internet of Things will advance the interaction with remote or virtual objects. Experts call this type of interaction “Tactile Internet”. The research team from Melbourne has made this vision a little bit more reality.
Their new mechanism will significantly optimize the interaction with machines. Their predictive touch response shortens network reaction times and makes interaction more dynamic than it is today.
“These response times impose a limit on how far apart humans and machines can be placed,” – Elaine Wong, University of Melbourne
In order to optimize the low reaction speed, the researchers are working with a learning algorithm that, thanks to machine learning, can predict the corresponding feedback.
Before the actual feedback is known, the module, which is called Event-based HAptic SAmple Forecast, or EHASAF for short, uses a neural network to predict the material that is touched. Once the material is detected, EHASAF adjusts its probability calculation and updates it to predict the appropriate feedback.
To test the functionality of their development, the team combined their module with VR gloves. When touching, for example, a virtual ball, the module went through feedback loops that were generated until it defined the actual material of the selected ball.
Although this development is still a very small step towards optimized human-machine interaction, it can pave the way to the tactile Internet.
Translated with www.DeepL.com/Translator (free version)
