Displayed here is Marvell's broad range of network switch solutions running ARM processors. The first board uses dual-core ARM Cortex-A9. has 24x10Gbps Ethernet ports and two additional 40Gbps ports. Marvell's ARM CPUs come embedded with their security engine that allows for monitoring data activity. We also see their prosumer range meant for professional use, with support for a 10Gigabit port.
EziSmart is a company based in Norway with factories in Shenzhen. The EziSmart K1 is a range of keyboard cases that one can insert a phone into, to enable a hardware (non-QWERTY) keyboard. Targeted towards the elderly, the case can be flipped close to reveal an SOS button and four buttons for instant use. The company's app displays the caller's name in the open window on the case, with hardware buttons for picking up and ending calls. Their feature phone comes with a heart rate monitor, a pedometer and keypad with large keys.
This wearable features a piece of glass that allows for a 800x480px display for visual information from the Android v5.1 interface. The wearable comes in B2B and B2C versions costing US$480 and US$639 respectively. Demonstrated is the consumer model that features a capacitive touchpad on the body that allows UI control and recognizes taps. Battery life is rated at 2-4 hours of moderate to heavy usage and over 4 hours of light usage. There's an 8MP camera, 32GB of storage, and a microphone integrated for voice commands.
Epson Movierio Pro is a combination of sensors and cameras to provide a VR-like experience. The Moverio is meant for industrial settings where information about their work could be displayed on the screen in real time. Marketed as a “smart headset”, Moverie Pro resembles Google Glass a lot but is a lot more limited in scope. It uses Android and has a battery life of 4 hours.
The Epson Movierio here is demonstrated through its applications; we see the use of the glasses in instantly producing translations of the words that are spoken to it. The headset is also capable of introducing depth to images like with augmented reality (AR) solutions.
Forum 8 shows their driving simulation on Oculus VR headsets, allowing people to drive around the streets of Tokyo using gaming setups, consisting of a steering wheel and a pedal box. The company’s main focus is to develop interactive 3D VR simulation and modeling software.
Sharp’s talking robot is on display, and it can respond to touch and voice. Company representatives demonstrate how it can make conversation, respond to questions about the weather, and even respond humanly when struck with a hand. The robot is not for sale and is only meant to showcase company capabilities.
inForce Computing, a Qualcomm partner, is focussed on utilising Snapdragon SoCs to power a multitude of applications – IoT, portable healthcare devices, robotics, and so on. At TechCrunch Disrupt, Keith Fleer, technical marketing manager at inForce, displays a robotics platform using a Snapdragon 615 SoC with peripherals such as a camera and gesture sensor, on a board with PWM output for servo motors. There is also a board using the Snapdragon 805 chipset (with a 2.7GHz quad-core CPU) that could be used to build video conferencing devices.
For more information see https://www.inforcecomputing.com/
The HardKernal ODroid-C2 is a board that outputs native 4K resolution over HDMI 2.0. It uses 4x ARM Cortex-A53 cores at 1.5GHz (Mali-450 GPU) coupled with 2GB DDR3 RAM. Upto 128GB eMMC HS400 and 200GB UHS-1 microSD cards are supported simultaneously. A Gigabit Ethernet port as well as 4USB Host and 1 USB OTG ports are present. The demo setup consists of a 4K TV playing 4K videos while using Ubuntu OS with MATE desktop environment. The board costs US$41.95 without the recommended eMMC memory.
An application engineer from NXP demonstrates here their S32V234 automotive-grade image recognition processor that is meant to be used in autonomous self-driving vehicle applications. The board uses 4 Cortex-A53 cores , and an "Apex" signal processor that allows the image to be split into parts and processed on with algorithms in addition to an ISP for filtering. The demonstration is a camera feed processed to show object tracking, with relative motion speed and direction indicated. The board is also meant to recognize traffic signals , and supports 4 cameras with specific framerates for 360° vision processing.