Development kits drive IoT design innovations

Development kits have become a mainstay of many component manufacturers’ toolkits when launching new parts, especially for microprocessors and microcontroller (MCU) makers. Many of them also collaborate with partners to deliver everything needed to help designers develop their prototypes. Dev kits typically bundle all of the hardware, software, and connectivity to fully evaluate the features and functionality of the new devices.

More recently, chipmakers have added machine-learning (ML) and artificial-intelligence functionality as adoption rates grow across applications and industries. In many cases, these development kits are created through partnerships.

Over the past several years, chipmakers have started to focus on development platforms to help engineers develop new solutions, including in the area of the internet of things. A few examples include STMicroelectronics’ Discovery kits or its SensorTile.box, a ready-to-use wireless IoT and wearable sensor platform kit for any skill level; Renesas Electronics Corp.’s Winning Combinations, though they are considered reference designs; and Texas Instruments’ (TI’s) LaunchPad development kits with BoosterPack plug-in modules.

Over the past year, chipmakers have developed a variety of development kits for a range of applications. Here is a sampling of those kits and platforms.

Renesas offers its Quick-Connect IoT modular development platform to simplify the prototyping of IoT systems. The platform consists of standardized boards and interfaces that enable designers to connect a wide range of sensors to their MCU development boards. The platform also includes core software building blocks that are portable between boards, helping to reduce coding requirements.

Renesas’s Quick-Connect IoT modular development platform (Source: Renesas Electronics Corp.)

As part of the platform, Renesas worked with Diligent, Inc. to create a new expanded I2C Pod Interface, Type 6A, for wider coverage. The company standardized on the new Pod 6A connector for new sensor Pods and MCU development kits, enabling designers to select the optimal combination for their prototype designs. MCU boards will have two Pod connectors: one for a sensor Pod and another for a communications module. In addition, the Pods can be cascaded for greater flexibility.

Renesas said that over 25 MCU development boards and kits are compatible with the new Pod Type 6A standard, either directly or through a small interposer board developed by Renesas.

One of TI’s newest LaunchPad Development Kits is based on the new Sitara AM2x MCUs. These devices help designers achieve up to 10× the computing capability of traditional, flash-based MCUs. They can be used in applications such as factory automation, robotics, automotive systems, and sustainable energy management.

The Sitara AM2x MCU portfolio, built on the ARM MCU cores, includes single and multicore devices running at speeds up to 1 GHz and integrates specialized peripherals and accelerators. The AM243x MCUs are the first family of devices available in the AM2x portfolio, featuring up to four ARM Cortex-R5F cores, each running up to 800 MHz.

The Sitara AM243x LaunchPad Development Kit, together with the Sitara MCU+ software framework, help developers start using the precision real-time control and out-of-the-box networking capabilities in the AM243x MCU. Developers also have access to application-specific reference examples, as well as other tools and software.

As ML/AI plays a bigger role in IoT applications, more component manufacturers are adding AI elements to their development kits. Knowles Corp. has recently introduced a couple of development kits that focus on edge AI. First is the AISonic IA8201 Raspberry Pi Development Kit that bundles hardware, add-on software, and algorithms to deliver voice, audio edge processing, and ML listening capabilities to IoT devices and systems across a range of industries. The kit gives OEM/ODM product designers and engineers a single tool to streamline design, development, debugging, and testing for prototyping devices for new use cases.

At the heart of the development kit is the ultra-low–power and high-performance AISonic IA8201 audio edge processor OpenDSP that meets a range of audio-processing requirements. The audio edge processor combines two Tensilica-based, audio-centric DSP cores: one for high-power compute and AI/ML applications and the other for very low-power, always-on processing of sensor inputs. The IA8201 has 1 MB of RAM on-chip that allows for high-bandwidth processing of advanced, always-on, contextually aware ML use cases and memory to support multiple algorithms.

Using the Knowles openDSP platform, the kit includes a library of on-board audio algorithms and AI/ML libraries. The kit also offers options for either two or three pre-integrated Knowles Everest microphones and includes two microphone array boards to help select the right algorithm configurations for the end application. Applications range from smart home and home entertainment to smart buildings, residential/commercial security, and inference engines for industrial and commercial. Support is available through the Knowles Solutions Portal for configuration tools, firmware, and algorithms that come standard with the kit.

Knowles AISonic IA8201 Raspberry Pi Development Kit (Source: Knowles Corp.)

Knowles’s AI-enabled true wireless stereo (TWS) development platform helps shorten product development from entry-level to premium applications. The TWS development kit includes pre-tuned and pre-configured earbuds designed by Knowles and paired with a Bluetooth-enabled processing platform.

Knowles worked with multiple partners to integrate premium features in the development kit that enable manufacturers to add features such as active noise cancellation, ambient mode, high-definition audio, voice command, voice-call algorithms, and AI-enabled conversation enhancement.

The processing platform is at the core of the TWS development kit, said Knowles, consisting of the Knowles IA8201 AISonic audio edge processor, optimized for advanced voice and audio processing, and Sony Semiconductor Solutions Corporation’s CXD 3781 Codec paired with a premium Bluetooth audio system-on-chip (SoC) to enable immersive experiences and noise filtering.

Enabling premium voice pickup, the kit also includes Alango Technologies’ Voice Communication Package with multi-microphone beamforming and its OnlyVoice technology, which combines external beamforming with internal sensor-based voice processing, which are both ported to the IA8201 processor.

Also included in the TWS kit is AITransparency+ from Chatable — the first AI specifically designed to leverage the next-generation AI processing capabilities of the IA8201 AISonic processor for “conversation enhancement.” AITransparency+ also features advanced, on-chip, proprietary deep-neural–network architectures performing over 100 million AI calculations per second. This enables selective acoustic enhancement of conversational speech signals without perceptible latency, said Knowles.

The development platform allows manufacturers to test their own configurations with detachable earbuds, integrate advanced features, and tune for performance and power consumption, said Knowles. The kit also allows for customization, enabling manufacturers to leverage pre-tuned algorithms from Knowles’s partners or tune their own algorithms for other use cases.

For edge IoT, QuickLogic Corp. developed an open-source SoC development kit that plugs into a USB port. The Qomu development kit, a small-form–factor ARM Cortex M4F MCU + embedded FPGA (eFPGA) combination, fits into a USB Type-A port, eliminating the need for hardware setup and cabling. Together with its small size, measuring about 9 × 13 mm, the Qomu board enables users to develop anywhere. The open-source SoC development kit targets edge IoT applications that require ultra-low power consumption and AI or ML capabilities. However, it can be used in a wide range of applications.

The Qomu development kit contains QuickLogic’s EOS S3 MCU + eFPGA SoC, providing a lot of processing power for a tiny development kit. The ARM Cortex-M4F MCU features up to 80-MHz operating frequency and up to 512-KB SRAM. The eFPGA features 2,400 effective logic cells, 64-Kb RAM, up to eight RAM/FIFO controllers, two dedicated multipliers to offload math-intensive functions, a 16-channel DMA, configurable SPI and I2C interfaces, and ultra-low power consumption (in microwatts).

The kit, which is optimized for the QuickLogic Open Reconfigurable Computing initiative, is supported by a variety of vendor-supported open-source development tools.

QuickLogic Qomu development kit (Source: QuickLogic Corp.)

Since the introduction of Analog Devices Inc.’s MAX78000 low-power neural-network accelerated MCU that enables AI in battery-powered IoT devices, the company (at the time Maxim Integrated Products, which was acquired by ADI in 2021) has partnered with a few companies over the past year to develop IoT-detection solutions. ADI teamed with Aizip to deliver low-power IoT person detection, using the MAX78000 and Aizip Visual Wake Words. The solution enabled the MAX78000 to detect people in an image using Aizip’s VWW model at 0.7 mJ of energy per inference, which is said to be 100× lower than conventional software solutions. It was developed with Aizip’s proprietary design automation tools to achieve greater than 85% human-presence accuracy.

ADI also partnered with Xailient Inc. to develop an IoT face-detection solution that detects and localizes faces in video and images at 12 ms per inference. The solution combines the MAX78000 with Xailient’s proprietary Detectum neural network, which optimizes the computational efficiency and low-power sleep modes provided by the MAX78000. Xailient’s neural network is said to draw 250× lower power (at 280 µJ) than conventional embedded solutions.

For developers, ADI offers the MAX78000EVKIT# evaluation kit. The hardware includes a digital microphone, a gyroscope/accelerometer, parallel camera module support, and a 3.5-inch touch-enabled color TFT display. It also includes a secondary display that is driven by a power accumulator for tracking device power consumption over time. GPIO as well as analog inputs are readily accessible through 0.1-inch pin headers. The primary system power as well as UART access is provided by a USB Micro-B connector. A USB-to-SPI bridge provides access to on-board memory, allowing large networks or images to load quickly, said Maxim.

For deep AI development, SmartCow, an AI engineering company specializing in video analytics and AIoT devices, recently introduced its new audio-visual Apollo development kit, designed for higher-level developers’ requirements. Built around Nvidia Jetson Xavier NX, Apollo enables developers to create applications with conversational AI capabilities.

The Apollo development kit features on-board visual and audio sensors, including four microphones, two speaker terminals, two 3.5-mm phone jacks, an 8-MP IMX179 camera module, and an OLED display. It also includes a 128-GB NVMe SSD for storage and comes pre-packaged with the Nvidia DeepStream and RIVA Embedded SDK toolkits. The six NLP examples demoing the kit’s capabilities include: text-independent speaker-recognition systems, speech-to-text and sentiment analysis, language translations and speaker diarization, and applications for abnormal sound and surveillance.

The development kits support two programmable buttons: a default with one-key recovery to help the developers ease device recovery and a programmable button that allows developers to add their own applications.

Also focused on ML/AI is a collaboration between SensiML Corp., a developer of AI tools for building intelligent IoT endpoints, and onsemi. Developing a ML solution for autonomous sensor data processing and predictive modeling, the collaboration combines SensiML’s Analytics Toolkit development software with onsemi’s RSL10 Sensor Development Kit to create a platform for edge-sensing applications.

The RSL10 Sensor Development Kit, featuring Bluetooth Low Energy connectivity, combines the RSL10 radio with a range of environmental and inertial motion sensors in a tiny-form–factor board that interfaces to the SensiML Toolkit. This helps developers add low-latency local AI predictive algorithms to their industrial wearables, robotics, process control, or predictive maintenance applications regardless of their expertise level in data science and AI, said SensiML. The auto-generated code enables smart sensing embedded endpoints that turn raw sensor data into critical insights so that appropriate action takes place in real time without the need for cloud analytics.

In some cases, semiconductor manufacturers will also create development kits with their distribution partners. One example is Avnet’s Edge AI Development Kit. It features Avnet Embedded’s SMARC computer-on-module, based on NXP’s i.MX 8M Plus applications processor, combined with a production-ready SimpleFlex Carrier and a 10.1-inch touch display, providing a high-performance computing solution for ML edge applications. It also includes a dual camera vision board that can support single or dual IAS camera modules based on onsemi image sensors.

Avnet said the new kit enables advanced AI and ML applications to run faster at the edge. It also helps designers to add new features such as face recognition, voice-command processing, and other compute-intensive ML algorithms with existing applications. The company also offers example applications that leverage the NXP i.MX 8M Plus NPU core, with 2.25 TOPS of performance.

Avnet Edge AI Development Kit (Source: Avnet)

Arrow Electronics Inc. also added to its IoT offerings with the recently released PSA Certified development kit for IoT devices, based on Infineon’s PSoC 64 secure microcontrollers. The PSoC 64 IoT Security Workshop Development Kit is a PSA Certified (Level 1) platform to help designers develop secure systems using the PSA Certified trusted methodology. It is Arrow’s first PSA Certified product.

Developed with Infineon, the kit includes the Infineon PSoC 64 Secure AWS IoT Pioneer Kit, Arrow PSoC 6 IoT sensor shield, Shield2Go kits, and AWS cloud enablement with certified functional APIs and integrated dashboard for monitoring and visualization. PSA Certified Level 2 PSoC 64 silicon offers a root of trust for trusted services such as crypto and secure storage. The development kit is a pre-tested and validated reference design.

Arrow Electronics PSoC 64 IoT Security Workshop Development Kit (Source: Arrow Electronics)

Helping IoT device developers meet industry and government standards for electronic security as well as with emerging IoT legislation, including NIST 8259A and EN 303 645, the kit has been evaluated by SGS Brightsight, a global independent security evaluation laboratory.