TILT

What it does

TILT is a smart wearable device for those who wants to interact more natural with digital world. It’s replace crowd of controlling devices(remote control) to one smart band.

Your inspiration

The idea of this project came up to us while we had working on other project – bionic prosthetic hand. The girl, we made it for, controlled her prosthesis by muscle sensors. In fact, prosthesis is the complicated robotic device. We came up with idea that we can use the same method of controlling devices and shift it to another various applications. Our world is crowed of different types of digital and electronic devices and its input devices or remote control. We were inspired of idea that we can create that gestured control device for routine use in our digital world.

How it works

TILT is a human-machine interface. It puts on user forearm and provides qualified interaction with a Virtual Reality, videogames, smartphones, Smart Houses and different types of robotic systems: copters, prosthesises or manipulators . The interaction bases on recognition hand movement and gestures. This makes the interaction easy and intuitive for users and involve them in process. Recognition system of the TILT is very sensitive. It recognizes the position and movements of hand in space. Also each of eight silver-plated sensors in an array around your forearm recognize even slight signal. Signal produces by muscles. When we make some gesture, muscles in our forearm act a certain way, called pattern. System of sensors notes and save that pattern during calibrating process and then recognizes the already familiar gesture. So you can even create your own gestures in the TILT-App.

Design process

Our aim was to design a compact, wireless device helps people gestured control various devices, and sets your hands free. The TILT development process contained 2 parts: the definition of an arm position; recognition of independent fingers and wrist movements. First we have chosen 9-axis sensor for accurate recognition of arm position. All large muscles involved in gestures are located in the forearm, so a method based on tracing muscular contractions is perfect for our aim: there is no need to place the sensors directly on the fingers or wrist. We started prototyping with one muscle sensor on the forearm - that one recognized any gestures like the same gestures. On the next stage we made our own muscle sensors to reduce muscle sensor size and improve functional parameters for our specific tasks. Tests showed us that positioning sensors over each of the muscle groups, improved recognition of independent finger movements. For the next prototype we made a circular array of sensors and developed software to trace signals from each sensor and recognize different gestures. We tested our prototype on people with various parameters and made sure that the muscle contractions patterns of certain gestures are the same. The TILT allows to use natural relaxed gestures to control devices.

How it is different

Most of modern human-computer interfaces and remote control devices have a large size, hamper the user's movement, or tie him to a space within which one can use it. Devices which recognize fingers and wrist movements doesn't allow to use slightest movements and natural gestures for controlling devices. First, the TILT has high functionality and recognition quality. The device allows you to recognize the fine motor skills of hands, which makes the controlling process easier and more natural. The device can be used for video games, VR, smartphone or computer control, for controlling drones, Smart Home systems or radio-controlled toys. It is easy and comfort to use. Hands are always free for routine. You can easily switch between controled devices with gestures. Control gestures do not require strong hand strain, they are relaxed and natural. The TILT is portable and wearable, it does not tie the user to the space and does not cause discomfort during use.

Future plans

Now we are creating a database on which recognition of complex gestures will be based. We are going to use neural network for gesture recognition. We are also working on new design improve comfort for users. After completing this tasks we plan manufacturing the TILT and launch it on market.

Awards

This project was awarded by FASIE in "UMNIK" 2017