Tag Archives: machine learning

New development of smart carpet and its applications

A floor sensor system was set up to detect gait activities by recording the amount of pressure applied in different places when someone walks across it.  

Picture: Mission: Impossible (1996) – Close Call Scene (5/9).

Fans of Tom Cruise may recall vividly in his movie Mission: Impossible (1996), Tom intruded to a room of cyber look through the ceiling. He was hovered just inches above the floor to gain access to a computer, being very careful not to touch the floor. 

Two decades later, we have smart carpet that would make Tom Cruise and his partners even more difficult to get their jobs done.

Smart carpet is generally referred to the floor finishes with embedded electronic devices that enable detection of movements that exert upon it. It is used for monitoring activities of elderly especially those who live alone.

Alongside the development of new behaviour biometric, smart carpet now becomes even more versatile. 

Research team at the University of Manchester

The prominent case here is the research team led by Prof. Krikor OZANYAN and Dr Patricia SCULLY at the University of Manchester. They adopted efficient machine learning methodology to classify gait data from raw tomography sensor data accurately without the need to reconstruct images.

A floor sensor system was set up to detect gait activities by recording the amount of pressure applied in different places when someone walks across it. The team did not use piezoelectric wires or pressure sensors / tiles as other researchers used to be. They used low cost standard polymer optical fibre that form grids to transmit and receive lights.

TrackingCarpet_m_0904.jpg.653x0_q80_crop-smart

Picture: Researchers demonstrate a piece of ‘smart carpet,’ which detects the footsteps of people who stand on it. (Photo: University of Manchester).

 

By using artificial intelligence, the changes in location and amount of lights as caused by human motions, e.g. walking or exercising, are recorded, plotted and analyzed. The complexity of the research lies not only on identifying the types of footstep patterns but also in the necessity to cope with the footstep spatial and temporal variances in the data captured from a real-world gait recording system.

The result was encouraging with high accuracy in identifying pre-defined variants of gait. 

Possible uses of smart carpet

The applications of magic carpet can be group into 2 categories: health and safety; and security. It can be installed to monitor old people’s movements in elderly home, as well as patients in clinics and hospitals, and thus becomes an integral part of the healthcare systems.

For security purpose, such a carpet can be placed to track and identify people who walked past on it at residential and office buildings. They can also be applied in areas where volume of people is high and constant surveillance is required to ensure public safety. These places include transport terminals (e.g. train stations and airports), museums, government premises, tourist attractions and places for entertainment and leisure.

[BEIT NEWS] 12 Oct 2018 – eNightLog for elderly, digital construction and cyber security at healthcare

polyu-logo

PolyU develops eNightLog System for caring Elderly with Dementia

The non-contact and non-invasive eNightLog system is embedded with event sequence tracking and different kinds of remote sensing and imaging technologies, based on innovative algorithm developed by the BME team of PolyU. This innovative technique has already been patented.

Seventeen systems have been installed and tested in a nursing home for 2 months this year for nighttime monitoring. During this period, 380 incidences of elderly leaving bed alone were recorded, with all them being successfully detected (100%), and only 2 times of false alarm occurred (0.5%). In addition, the system recorded 525 events of caregivers visiting elderly and accompanying the elderly leaving, and accuracy rate is 100%. 

Keywords: eNightLogProf. ZHENG Yongping; The Hong Kong Polytechnic University; dementia; Alzheimer’s disease

2-700

Picture: eNightLog – the research team led by Ir Prof. ZHENG Yongping in the Hong Kong Polytechnic University.


download (1)

Imerso offers intelligent technology through 3D capture and BIM

 

Working on Imerso is simple as Users start by loading the software with all the data from the planning stage, which includes BIM, geometry, schedules as well as tolerance parameters and then these data are used as monitoring targets. When, site works continues, users upload 3D data on field. The next step perform by Imerso is to register that 3D data directly to the 3D building information model and displays the result, through this it is easy for the users to compare the current on-site conditions against the original design intent.

Keywords: Imerso; 3D; Building Information Modelling (BIM); digital construction; Frederico VALENTE

imerso

Picture: Imerso’s digital platform for construction.


 

sensors-online_logo_

 

… by leveraging machine learning, organizations can largely diminish the security concerns that come with Internet of Things (IoT). Machine learning enables data exchanges to be monitored within the organization, and with external parties to detect anomalies that are not considered the norm. Machine learning can also aid in predicting threats, in its ability to analyze historical data from specific trends, which can be evaluated from the big data produced by the algorithms.

Keywords: Healthcare; Internet of Things (IoT); cyber security; machine learning; behavior analytics 

HERO

Source: https://www.sensorsmag.com/iot-wireless/how-to-combat-security-risks-healthcare-iot-security

[BEIT NEWS] 21 Sept 2018 – Landscape Design Visualizer and MIT’s AI to sense people

1

Software Targets Efficient Landscape Design

 

The Landscape Design Visualizer helps users to imagine how their landscape design ideas will look in real-world applications. “Industry leaders typically embrace our tools first, and Oaks Pavers is a leader in the manufacturing of various landscape products,” said Dan Dempsey, President of Chameleon Power. “Visualization is quickly becoming a standard in landscape design, and we are happy to provide this experience to the Oaks Pavers customers.” 

Keywords: Oaks PaversLandscape Design Visualizer; Chameleon Power; landscape design

Source: Chameleon Power’s visualizing technology


 

logo_192

MIT using radio waves to monitor patients’ health through walls

 

A professor at the Massachusetts Institute of Technology has developed a device that can monitor your health using radio signals and machine learning, even through walls.

The gadget looks like a box and works like a Wi-Fi router; the low-power wireless signals reflect off nearby people’s bodies and carry information back to the box, the MIT tech review said. The box analyzes the information and offers up health data. The device senses your movements, even if you’re just breathing, by recording changes in the electromagnetic field around you.

Keywords: Massachusetts Institute of Technology (MIT); Dina Katabi; machine learning; artificial intelligence; smart home

Picture:  AI Senses People Through Walls