PORTLAND, Ore. — The former director of Apple's Siri is taking Samsung's version of the artificial intelligence system to the next level. Luc Julia, vice president and innovation fellow at Samsung's Open Innovation Center in Menlo Park, Calif., demonstrated SAMI (Samsung Architecture for Multimodal Interactions), the Siri-like system central to Samsung's Internet of Things (IoT) strategy, at the MEMS Executive Congress 2013 in Napa, Calif., Nov. 7-8.
Luc Julia heads the team working on SAMI, which is being developed at the center. SAMI is an interactive artificial intelligence (AI) similar to Apple's Siri that Julia helped develop when at Apple. Samsung's SAMI, however, goes far beyond Apple's Siri by aggregating sensor data from all types and brands of IoT devices in the cloud. The open system will then allow Samsung ecosystem partners -- some financed by a $100 million accelerator fund -- to perform deep analytics on that data before sending smart advice back to users.
"IoT is all about sensing and collecting the data -- and then transporting it [to the cloud]. Once it's there, it's about massaging it and sending it back to the user," said Julia, who described the sensing and collecting stage with the "big data" moniker, but who spoke of the machine learning and analytic engines sending data back to the user as "small data" -- advice that is directly relevant to people's lives.
Not just for astronauts anymore: IoT wearables monitor all a person's vital signs and transmit them to the cloud for analysis.
Samsung's Open Innovation Center in Menlo Park is one of 27 OICs worldwide, and SAMI is just one of several ongoing projects there. So far, Samsung's Menlo Park OIC is exploring possibilities with about 48 companies regarding providing the innovative sensors and actuators for wellness, smart-homes, and smart-cars, as well as the cloud-based "brains" for projects like SAMI. Samsung is engaging startups that it will fund with early-stage investments in the $100,000 to $2 million range from Samsung's $100 million accelerator fund.
According to Julia, who left Apple last year, there are about 20 billion IoT devices out there today, but he predicts that number will grow to 1.5 trillion by 2020. Driving that growth will be what he calls the "explosion" of the smartphone, whereby the sensors inside it, and dozens more in dedicated IoT devices, explode like shrapnel that becomes embedded into wearable devices around the body.
At the MEMS Executive Congress he illustrated the point by wearing a half-dozen sensors himself -- including the Basis smartwatch, Fitbit activity monitor, and a Vitalconnect electrocardiogram monitor. SAMI was able to aggregate all the data streaming from those sensors, integrating it with weight measurements from a Withings scale and food intake information from a HapiLabs smart-fork. When Julia verbally asked SAMI "How am I doing?" it displayed a common dashboard showing readouts from the sensors, as well as provideing verbal advice, such as suggesting that he needed to lose the four pounds he had put on recently.
Samsung's Architecture for Multimodal Interactions: SAMI (green) aggregates data from sensors (top), filters and augments it before storing it in a normalized format, then analyzes it to supply coaching tips to users, which can be accessed with an application programming interface by companies like Lark, or can go directly to an Android device like the Pebble smartwatch.
SAMI is an internal project that is only about six months old, but eventually will be made into an open-system platform to which Samsung's ecosystem partners can freely contribute to transform the user experience.
"We want the user experience to be magical, to be of things that are not expected," said Julia.
Samsung Vice President of Strategy at its Open Innovation Center (Menlo Park, Calif.) described its Internet of Things (IoT) strategy. (SOURCE: R. Colin Johnson)
Today data is isolated in silos, according to Julia, but SAMI aims to break down the silo, allowing smart applications to access cross-silo data sets. To do so, SAMI will aggregate data in any format -- structured or unstructured -- then provide its partners with a common method of accessing any of it.
"We tell startups just to give us their data in the format they already have, and what we do is create a normalized a way of accessing it," said Julia "Security is also very important; we need to control who has access."
Julia cited Lark Technologies -- makers of an exercise, diet, and sleep coach -- as one of the companies providing an exemplary SAMI-like service that analyzes data from sensors, then coaches the user to improve his or her life.
— R. Colin Johnson, Advanced Technology Editor, EE Times