Proving the Business Case for the Internet of Things

AI mattress automatically adjusts to body shapes

Steve Rogerson
March 13, 2018

A mattress developed by Silicon Valley-based iFutureLab can autonomously recognise and adapt to individual body shapes and postures in real time to offer better support and comfort.
The autonomous adaptation functionality of AI Trackbot is driven by artificial intelligence (AI), and the technology has recently entered into mass production. IFutureLab plans to apply this technology to furniture, with mattresses as the initial focus.
Sleep occupies one third of a human life. Diseases, poor concentration and low productivity caused by sleep problems have resulted in economic losses in the trillions of dollars every year. According to research at the iFutureLab Sleep Medicine Center (SMC), support has an influence on the health of one's skeletal and muscular systems and blood circulation during sleep. Over 90% of sleep problems are closely linked to improper support.
According to a recent report by the UN Desa, the world population reached nearly 7.5 billion, which as a whole require approximately 160 million different variations of support during sleep. Modern medical research indicates that people of varying ages, heights, weights and body shapes have different support requirements. In addition, variable health conditions and sleep preferences will affect one's quality of sleep. Further, as people toss and turn throughout the night, sometimes in excess of ten times, the support required for each position will be different.
All these factors have created a major challenge for traditional mattresses to meet the various changing support requirements. Traditional mattresses use inner springs, foam and air cushions for support, which each have their own inherent weaknesses. The limited materials cannot meet individual support needs of different people, and the fixed structures of traditional mattresses cannot adapt to changing sleep positions in real time. As a result, traditional mattresses fail to provide suitable support for the best sleep experience.
While compared with traditional mattresses, such as innerspring mattresses, latex mattresses, air beds and electric beds, the major advantages of an AI mattress include autonomous recognition of individual body shape and sleep position, calculating optimal support for each part of the body based on sleep medicine databases, and autonomous adjustment of height and firmness for changing positions. Such functions ensure that the body stays properly aligned and properly supported during sleep.
Composed of an AI sensing system, autonomous adjustment system, TensorFlow deep learning network, central processing system and sleep medicine databases, AI Trackbot has high detection, analysis and processing capabilities. Powered by the sleep medicine databases and AI technology, it can independently recognise 160 million sleep models and calculate the optimal pressure distribution for every part of the body within 0.01 second.
The system collects real-time data to build an exclusive model for the user who lies on it through ergonomics based on his or her real-time pressure distribution and age, height, weight, body shape and sleep preferences. Optimal postural support and pressure relief for the body are achieved by adjusting height and firmness along with the changing sleep positions.
"When lying on a mattress equipped with AI Trackbot, I can feel the slow movements and tender adjustments under my body no matter if I'm on my back or side,” said Albert, an early tester of the system. “It's hard to explain exactly how I feel, but it's really very relaxing and comfortable. I definitely enjoyed my experience."
More than 30,000 experiments conducted by iFutureLab show that it can effectively reduce body movements, tossing and turning, the probability of waking up at night, and the percentage of discomfort in muscle and spinal regions upon waking up in the morning. This in turn increases the percentage of deep sleep over a night's rest. At the same time, through continuous follow-up studies, iFutureLab has found that users with chronic pain in the neck, shoulder, waist and leg areas and with pre-existing problems such as nerve compression or poor breathing have been relieved of such issues after using the mattress for a period of time.
The company has already begun cooperation with some furniture and mattress enterprises.
Heka is a membership-based company in Delaware that designs and crafts customised high-end furniture for its members. As a strategic partner of iFutureLab, Heka launched the AI mattress last month. Rogan Smith, CMO of Heka, believes it can help provide the ultimate sleep experience for its members.
Customisation of the Heka bed has already begun globally. The firm has received orders from many celebrities and elites, including members of royal families, founders and chairmen of Fortune Global 500 and multinational corporations.
"This technology will also help those millions of athletes going through injuries and chronic pain sufferers to get the pain relief they need,” said Kent Scott, director of the iFutureLab AI Trackbot section. “They just need to simply input their individualised chronic pain information or sleep preferences into the app. AI Trackbot will then calculate the optimal support under such situations, allowing for the reduction of the impact of these physical problems on sleep quality. In addition, AI Trackbot is also cooperating closely with many hotels, enabling people with different body types and sleep preferences to truly obtain the best night of sleep while staying in a hotel."
A total of 43 scientists and engineers have developed the technology. Analysis on the relationship between support and healthy sleep has been made based on ergonomics, cardiovascular medicine and musculoskeletal medicine. After five years of efforts, iFutureLab SMC has accomplished the test and observation of total 2.3 million sets of experimental data involving subjects from 13 countries or regions and 37 nationalities.