Around The Home

Pressure-sensing GravitySpace floor adds new dimension to the smart home

Pressure-sensing GravitySpace floor adds new dimension to the smart home
GravitySpace recognizes people and objects based on the pressure imprints they leave on the floor
GravitySpace recognizes people and objects based on the pressure imprints they leave on the floor
View 4 Images
The GravitySpace prototype senses pressure at 1 mm resolution and projects across an active area of 8 m² in a single seamless piece
1/4
The GravitySpace prototype senses pressure at 1 mm resolution and projects across an active area of 8 m² in a single seamless piece
GravitySpace recognizes people and objects based on the pressure imprints they leave on the floor
2/4
GravitySpace recognizes people and objects based on the pressure imprints they leave on the floor
GravitySpace derives the foot location above the floor based on pressure distributions of the other foot
3/4
GravitySpace derives the foot location above the floor based on pressure distributions of the other foot
GravitySpace sees the scene from Figure 2 as a set of contacts (circles, lines, and text added for clarity).
4/4
GravitySpace sees the scene from Figure 2 as a set of contacts (circles, lines, and text added for clarity).
View gallery - 4 images

Smart floors could soon be part of our smart homes. Scientists in Germany have developed a high-resolution pressure-sensitive floor that can accurately keep track of people and furniture in rooms. Dubbed "Gravity Space," the floor can detect poses, movements and collisions and create a mirror-like inverse projection of the goings-on above. The technology could have a wide range of applications ranging from home security and automation to interactive gaming.

The team from the Hasso Plattner Institute in Potsdam has created an eight square meter back-projected floor prototype, a set of touch-sensitive furniture and algorithms that identify users, furniture and poses above the ground based on their textures.

The smartfloor uses what the researchers call a multitoe infrastructure, which relies on per-pixel pressure to enable the floor to locate and analyze users’ soles and recognize foot positions.

The GravitySpace prototype senses pressure at 1 mm resolution and projects across an active area of 8 m² in a single seamless piece
The GravitySpace prototype senses pressure at 1 mm resolution and projects across an active area of 8 m² in a single seamless piece

The Gravity Space prototype is made up of a slab of 6.4 cm (2.5 in.) thick glass installed in a hole cut into a standard floor. There’s a room below with an infrared camera and high-resolution video projector that tracks footprints and beams video up onto the glass. Infrared LEDs surround the flooring coated with a rubbery, pressure-sensitive film. A footstep on the surface makes the film interfere with the infrared light, creating an image of the footprint captured by the camera in the room below. The system can identify users based on their soles, track users' foot and body postures, and enable high-precision interaction by focusing on their feet.

The scientists say the system has several advantages over camera-based tracking. First, it provides a consistent coverage of the room wall-to-wall. Secondly, it is less susceptible to people in the room blocking vision. It also allows for simple algorithms and, finally, it is less intrusive on privacy.

Gravity Space has numerous potential applications. In addition to home and office security, the technology could be used help take care of the elderly where, for example, a caregiver would be alerted when the patient falls. It could even be used for pure fun – think kicking a CGI football around a room.

IBM is also awake to the potential of smart floors, having patented a multi-touch smart floor that can detect home intruders or call emergency services like 911 when someone has fallen down or suffered a heart attack.

The Hasso Plattner Institute team members are Alan Branzel, Daniel Hoffmann, Marius Knaust, Patrick Lühne, René Meusel, and Stephan Richter, supervised by Christian Holz, Dominik Schmidt, and Patrick Baudisch at the Human Computer Interaction Lab.

The project has been developed in collaboration with Shahram Izadi, Steve Hodges, and Scarlet Schwiderski-Grosche at Microsoft Research in Cambridge, UK.

Source: Hasso Plattner Institute via New Scientist

View gallery - 4 images
3 comments
3 comments
piperTom
Is it April 1st already?
The "mirror-like inverse projection" does mimic a mirror... from one, very limited point of view. That same image is going to look very weird from every other direction.
Stephen N Russell
Must for theme parks, Univs, colleges, acting schools, Education & for future interior planning.
Philipp Koschinski
There are some Problems, for example, if two people wear the same shoes or similar shoes, the computer recognizes them as the same person. (Security aspect?) Also, don't ever change shoes -- or you wouldn't be recognized anymore.