NXP Semiconductors has unveiled an MCU based voice control solution qualified with Amazon’s Alexa Voice Service (AVS). This enables original equipment manufacturers (OEMs) to quickly, easily and inexpensively add voice control to their products, giving their customers access to rich voice experiences with Alexa. Built on an NXP i.MX RT crossover platform, this MCU-based AVS solution enables low latency, far-field, “wake word” detection; embeds all necessary digital signal processing capabilities; runs on Amazon FreeRTOS; and includes an Alexa client application. This MCU-based AVS solution provides OEMs with a self-contained, turnkey offering that enables them to quickly add Alexa to their products. It includes the MCU, the TFA9894D smart audio amplifier, optional A71CH secure element and comes with fully integrated software. It also features noise suppression, echo cancellation, beam forming and barge-in capabilities that enable use in acoustically difficult environments.
NXP offers at its Mougins, Sophia-Antipolis facilities a product testing service for Alexa Built-in products, available to its customers desiring to test their devices before submitting to Amazon for final evaluation. If a customer’s product supports music and/or is far-field enabled and uses a “wake word” to initiate interactions with Alexa, additional testing is required prior to submitting products to Amazon for evaluation. This is where Pro-Support Audio Voice Services helps to complete the self-test checklists.
Estone is launching an “EMB-2238” Pico-ITX board for audio and voice control applications that runs Linux on an i.MX8M and offers a dual-DSP audio hub and DAC, 40-pin GPIO, and optional PoE and second GbE.
Toledo, Ohio based Estone Technology (known for its former Habey brand) offers a variety of Linux-friendly Pico-ITX boards, including boards based on the i.MX6 (EMB-2230), i.MX6 UL (EMB2200) models, and Intel Cherry Trail EMB-2610. The company recently announced (via Electronics Weekly) an EMB-2238 board with the same 100 x 72mm form factor. The SBC builds on the audio strengths of NXP’s i.MX8M SoC with the help of high-end audio circuitry from Cirrus Logic.
(click images to enlarge)
The EMB-2238 uses the quad-core version of the 1.5GHz, Cortex-A53 equipped i.MX8M, which also includes a GPU and 266MHz Cortex-M4 chip. Estone provides a Yocto Project stack based on Linux kernel 4.9, Qt, and Wayland. It also supports Android 8.1.0.
Other i.MX8M Pico-ITX boards we’ve seen include Kontron’s dual-GbE pITX-iMX8Mand F&S Elektronik Systeme’s up to 8GB LPDDR4 armStone MX8M. There’s also a larger, 136.7 x 87mm Nitrogen8M SBC from Boundary Devices.
All these boards tap the i.MX8M’s extensive digital audio skills to varying degrees, but the EMB-2238 is even more focused on audio and voice control applications. It adds a Cirrus Logic CS47L24 smart codec module with a dual-core, 300-MIPS DSP and audio hub. The triple-DAC device offers a 115 dB dynamic range, an 8-192kHz sample rate, and Enhanced DRE processing (eDRE) for 121dB SNR.
EMB-2238 (left) and Cirrus Logic CS47L24 audio module block diagram
(click images to enlarge)
The CS47L24 drives the EMB-2238’s dual digital MEMS microphone header, which features multi-mic noise suppression and acoustic echo cancellation (AEC). A 40-pin expansion header provides omni-directional, spatial 8-channel digital audio/DMIC inputs (SAI5) for the mic array, among other I/O including PCIe. Additional audio features on the SBC include a Class D, 2W mono speaker, an 8-channel digital input and output (SAI1), and SPDIF and QSPI audio interfaces.
The EMB-2238 ships with the Amazon AVS (Alexa Voice Service) Device SDK, as well as Sensory’s TrulyHandsfree wake word engine. It also supports the Snips AI voice control assistant, including support for off-line operation (see video demo farther below).
You can purchase the SBC with 2GB to 4GB LPDDR4, and a microSD slot and 8GB iNAND are also available. For communications there’s a WiFi/BT module and a GbE port with optional an Power-over-Ethernet (IEEE 802.3af) or PoE+ (802.3at) module that can also power an attached LCD panel. A separate option provides a second GbE port via a PCIe add-on card that also integrates a 9-36V DC input, GPIO, an ambient sensor, and an LED control for light bars.
Optional PoE (left) and GbE add-ons
(click images to enlarge)
The EMB-2238 is equipped with a 4K-ready micro-HDMI port and HD-ready MIPI-DSI with optional 10.1-inch touch-panel. Other features include a MIPI-CSI camera interface, USB 3.0 OTG Type-C port, dual USB 2.0 host ports, and 2x internal USB interfaces.
The SBC provides a RS-232/RS-485 terminal block, RS-232 header, and the 40-pin header. A 5V DC header offers an alternative to the optional PoE and 9-36V input. The board also provides a watchdog and 0 to 60°C support. As usual with Estone, you get comprehensive documentation.
At the CES show last month, Mercedes-Benz its NVIDIA-powered MBUX infotainment system–a next-gen car cabin experience can learn and adapt to driver and passenger preferences, thanks to artificial intelligence.
According to NVIDIA, all the key MBUX systems are built together with NVIDIA, and they’re all powered by NVIDIA. The announcement comes a year after Huang joined Mercedes-Benz execs on stage at CES 2017 and said that their companies were collaborating on an AI car that would be ready in 2018.
Powered by NVIDIA graphics and deep learning technologies, the Mercedes-Benz User Experience, or MBUX, has been designed to deliver beautiful new 3D touch-screen displays. It can be controlled with a new voice-activated assistant that can be summoned with the phrase “Hey, Mercedes. It’s an intelligent learning system that adapts to the requirements of customers, remembering such details as the seat and steering wheel settings, lights and other comfort features.
The MBUX announcement highlights the importance of AI to next-generation infotainment systems inside the car, even as automakers are racing put AI to work to help vehicles navigate the world around them autonomously. The new infotainment system aims to use AI to adapt itself to drivers and passengers— automatically suggesting your favorite music for your drive home, or offering directions to a favorite restaurant at dinner time. It’s also one that will benefit from “over-the-air” updates delivering new features and capabilities.
Debuting in this month (February) in the new Mercedes-Benz A-Class, MBUX will power dramatic wide-screen displays that provide navigation, infotainment and other capabilities, touch-control buttons on the car’s steering wheel, as well as an intelligent assistant that can be summoned with a voice command. It’s an interface that can change its look to reflect the driver’s mood—whether they’re seeking serenity or excitement—and understand the way a user talks.
To address the convergence of immersive sensory experiences fueled by voice, video and audio demands, NXP Semiconductors has launched the i.MX 8M family of applications processors. The processors combine robust media capabilities on one chip. Voice commands are expected to dominate 50% of all searches in the next two years, increasingly thinner TVs are driving the popularity of sound bars for home automation, and consumers are embracing the IoT for creating more convenient richer sensory-driven experiences.
The NXP i.MX 8M processors address designers’ requirements for one platform that combines A/V and machine learning to create connected products that can be controlled via voice command. The chips provide the process technology and edge computing needs to manage and reduce the command and question response time of smart connected devices. The i.MX 8MF is suited for a wide range of residential IoT and device control applications including everything.from smart TVs, television subscription services, sound bars and other smart speakers, to streaming media players and DVR/PVR. The processor family is also ideal for managing lighting, thermostats, door locks, home security, smart sprinklers, other systems and devices for a more intuitive and responsive home environment.
NXP’s i.MX 8M family’s features that include:
Video and audio capabilities with full 4K Ultra HD resolution, High Dynamic Range (HDR) and the highest levels of pro-audio fidelity
Performance and versatility with up to four 1.5 GHz ARM Cortex-A53 cores, flexible memory options, and high-speed interfaces for flexible connectivity
Advanced Human Machine Interface (HMI) featuring dual displays, vision procession unit (VPU), and an enriched user experience
Infineon has made a strategic minority investment in XMOS Limited, a Bristol based fabless semiconductor company that provides voice processors for IoT devices. Infineon leads the recent $15 million Series-E funding round. According to Infineon, cars, homes, industrial plants and consumer devices are rapidly becoming connected to the Internet: 3 years from now, 30 billion devices will belong to the IoT. While today the interaction between humans and machines is mostly done by touch, the next evolutionary step of IoT will lead to the omni-presence of high-performance voice control. Infineon Technologies wants to further develop its capabilities to shape this market segment.
Today, voice controllers, used in voice recognition systems, struggle to differentiate between speech from a person in the room, and a synthesized source such as a radio, TV; they often identify the voice of interest based on the loudest noise. Earlier in 2017 Infineon and XMOS demonstrated an enhanced solution to overcome these issues, using intelligent human-sensing microphones and gesture recognition. The solution featured a combination of Infineon’s radar and silicon microphone sensors to detect the position and the distance of the speaker from the microphones, with XMOS far field voice processing technology used to capture speech.