CEVA Debuts Integrated Hardware and Software Platform for Contextually-aware IoT Devices
SenslinQ combines integrated software with sensors, Arm- or RISC-V-based MCUs, Ceva's DSP units, and other connected technologies for contextually-aware IoT devices.
CEVA, Inc. has announced SenslinQ, an integrated hardware IP and software platform designed to streamline the deployment of contextually-aware IoT products by marrying integrated software with sensors, audio components, and other connected technologies.
This technology is built to enhance IoT functionality as smart technology becomes omnipresent. According to CEVA, IoT contextual awareness is a key feature of “many devices such as smartphones, laptops, AR/VR headsets, robots, hearables, and wearables.”
SenslinQ is built to unify IoT device components to enhance the user experience. Image (modified) used courtesy of CEVA, Inc.
CEVA unveiled SenslinQ at CES 2020—which has included several themes relevant to hardware engineers. The SenslinQ system highlights the growing importance of harmonious hardware-software interaction within our modern electronics.
What's Inside SenslinQ?
CEVA says SenslinQ's hardware architecture is comprised of three main pillars: a microcontroller, a proprietary CEVA digital signal processor (DSP), and a wireless connectivity interface.
Arm- or RISC-V-based MCU
The microcontroller is either Arm- or RISC-V-based. Some digging into CEVA’s Bluetooth stack uncovers compatibility with the Arm Cortex-M series, ARC EM, RISC-V, Cortus APS, and AndesCore.
Block diagram of SenslinQ's hardware platform. Image used courtesy of CEVA, Inc.
Because many smart devices feature these processors, SenslinQ may be a flexible option for such IoT devices.
Adjacent to the processor is the CEVA-BX DSP. This signal processor stabilizes and analyzes incoming information from a device’s array of sensors. This input can include audio waves, voice control, temperature readings, pressure, and more. The DSP also reduces interference (noise) while boosting receiver sensitivity as needed.
Have you ever asked your Amazon Echo or Google Home a question?
The digital signal processor determines whether your voice is human or noise before processing requests. For the BX and BX2, CEVA’s components power neural networking and AI processing. This aids the device in speech recognition.
Wireless Connectivity Interface
SenslinQ's hardware offerings are powered by customized software stacks. The wireless connectivity interface—that third pillar—includes Bluetooth and connectivity IPs, plus a Wi-Fi chip.
Block diagram of SenslinQ's software platform. Image (modified) used courtesy of CEVA, Inc.
These IP cores are said to grant each SenslinQ device its functionality, logic, and overall internal layout. This optimization is key when dealing with compact devices.
CEVA's Own Bluetooth and Connectivity Platforms
CEVA offers its own RivieraWaves Bluetooth platform. This solution incorporates embedded Bluetooth Low Energy (BLE) and dual modes into a unified system on a chip (SoC). It can also utilize a specialized integrated circuit (IC).
Diagram of RivieraWaves Bluetooth platform. Image used courtesy of CEVA, Inc.
Both configurations include a baseband controller and a combined modem/RF unit—the latter being compatible with chips from multiple providers. This “processor agnostic” approach allows device manufacturers to choose preferred RF chips.
CEVA’s Dragonfly NB2 platform is a multi-application connectivity solution. Like RivieraWaves, it includes a processor, baseband, and RF transceiver. These components pair well with sensors found within IoT devices.
Block diagram of CEVA-Dragonfly NB2. Image (modified) used courtesy of CEVA, Inc.
By contrast, Dragonfly uses the CEVA-X1 processor. The X1 enables cellular and short-range communication; it also allows “[global] positioning, always-on functionality, and speech processing.”
CEVA asserts its optimizations allow devices to run on a single battery for 5 to 10 years. The company also claims that its 3.6 CoreMark/MHz benchmark bests other competing options with a single core.
General Improvements and Future Outlook
The SenslinQ aims to centralize processes that traditionally require a bevy of supplemental components. SenslinQ devices are said to analyze their environment in real-time using coordinated sensor readings.
Because these environments are dynamic, smart devices with SenslinQ are designed to learn as conditions change. All incoming data is filtered during processing, which in theory will boost accuracy.
This also offers readings contextual importance. SenslinQ allows sensors to work together instead of drawing conclusions from singular readings. Bundled software allows devices to differentiate between voices, sounds, activities, and proximity. Multiple IoT devices—and their users—may benefit from this awareness.
Connected devices are also cloud-enabled. SenslinQ’s communications stacks allow data transmission to computers and other personal devices.
It's possible that we may see this technology both in the home and workplace as it matures.