Table of Contents
1. NVIDIA Jetson Nano.
A small, powerful computer that lets one run multiple neural networks in parallel for applications like image classification, object detection, segmentation, and speech processing, the NVIDIA Jetson Nano is rather easy-to-use and it takes up as little as 5 watts. This developer kit can be powered by micro-USB and is also supported by NVIDIA JetPack, which includes a board support package (BSP), Linux OS, NVIDIA CUDA®, cuDNN, and TensorRT™ software libraries for deep learning, computer vision, GPU computing, multimedia processing, and much more. The software, on the other hand, is available using an easy-to-flash SD card image, making it easy to get started. The best part is that the same JetPack SDK is used across the entire NVIDIA Jetson family of products and thus allows for full compatibility with NVIDIA’s world-leading AI platform for training and deploying AI software. This in turn reduces complexity and work for developers.
Related post: AN810-XNX: Aetina’s new nVidia Jetson-based platform
One of the biggest advantages is that it is an AI computing platform offering GPU-accelerated parallel processing. The Jetson Nano has a 128 CUDA core GPU based on the Maxwell architecture.
Also, Nvidia has an open source project called Jetson Inference; it runs on all its Jetson platforms, including the Nano. The Jetson Interference demonstrates various machine learning techniques such as object recognition and object detection. This makes the Nano an ideal starting point for developers looking to build real-world machine learning projects.
While the hardware is very impressive, better support is needed on the software end. Another deterrent is the cost. Priced at $99, it is way more expensive than its rivals such as the Raspberry Pi.
NVIDIA Jetson Nano projects: The Jetson Nano is being used in the development of scores of new small, low-power Artificial Intelligence (AI) systems. It is also popular when it comes to embedded IoT applications, including entry-level Network Video Recorders (NVRs), home robots, and intelligent gateways with full analytics capabilities.
2. Raspberry Pi 4 .
The all-new Raspberry Pi 4 Model B is the latest version of the low-cost Raspberry Pi computer. This credit-card sized electronic board comes with several improvements. For one, it has a USB-C for the power connector and this USB can support an extra 500mA of current, thus offering1.2A for downstream USB devices. The type-A (full-size) HDMI connector has been replaced with a pair of type-D (micro) HDMI connectors; this in turn allows for dual display output within the existing board footprint. The Gigabit Ethernet magjack is now to the top right of the board, instead of the bottom right in Raspberry Pi 3.
In addition to this, the Raspberry Pi 4 boasts an overhauled operating system, based on the forthcoming Debian 10 Buster release. The user interface has been modernized and updated applications include the Chromium 74 web browser. Also, the legacy graphics driver stack used on previous models has given way to the Mesa“V3D” driver which in turn allows for doing away with almost half of the lines of closed-source code in the platform, in addition to offering the ability to run 3D applications in a window under X, OpenGL-accelerated web browsing and desktop composition.
As mentioned above, the Raspberry Pi 4 is faster and more capable than its predecessors due to the scores of improvements. It can also play 4K video at 60 frames per second, unlike Raspberry Pi 3. Other Pi 4 features include built-in Wi-Fi and Bluetooth. The latest board is also capable of booting directly from a USB-attached hard drive or pen drive.
The Raspberry Pi 4 can run Windows 10 but it’s important to keep in mind that it is not anything like the full desktop version of Windows 10; it is but a low cost alternative. Similarly, while it can run Windows desktop apps, it will need a lot of effort to make that happen, and even then apps are more likely to run poorly. It also continues to lack internal storage but one can use SD cards to address this.
Raspberry Pi 4 projects
Like its predecessors, the Raspberry Pi 4 is an incredibly useful IoT board that can be used for numerous projects. Examples include filming one’s own stop motion video, building one’s own Pi web server, a Raspberry Pi home security system, or a Raspberry Pi home automation system, building a virtual jukebox, creating a network monitoring tool, a Raspberry Pi robot and many more.
A low-cost, low-power system from a chip (SoC) series that has been created by Espressif Systems, ESP32 comes with Wi-Fi & dual-mode Bluetooth capabilities. One of the key features is its dual-core or single-core Tensilica Xtensa LX6 microprocessor with a clock rate of up to 240 MHz. Highly integrated with built-in antenna switches, RF balun, power amplifier, low-noise receive amplifier, filters, power management modules, touch sensitive pins, built-in hall effect sensor and temperature sensor, ESP32 has been engineered for mobile devices, wearable electronics, and IoT applications.
The successor to ESP8266, ESP32 too is rather easy on the pocket. Moreover, it offers ultra-low power consumption thanks to its power saving features such as fine resolution clock gating, multiple power modes, and dynamic power scaling. The ESP32 also has many more GPIOs than the ESP8266.
Priced at between $6 to $12, the ESP32 costs slightly more than the ESP8266, which costs between $4 to $6.
Given its many features, the ESP32 can be used for a slew of DIY (Do It Yourself) IoT projects and DIY smart home projects. Some instances include sensor-based projects such as creating an all-in-one ESP32 weather station shield and working with barometric sensor; data logging projects like recording the temperature to MicroSD card and web-based projects such as setting an ESP32 Access Point (AP) for the web server.
ESP8266 is a low-cost Wi-Fi microchip with full TCP/IP stack and microcontroller capability produced by Shanghai-based Chinese manufacturer Espressif Systems. It contains a built-in 32-bit low-power CPU, ROM and RAM. It is a complete and self-contained Wi-Fi network solution that can carry software applications as a stand-alone device or connected with a microcontroller (MCU). The module comes fitted with an AT Command firmware that can be used with any MCU via COM port. Fitted with a L106 32-bit RISC microprocessor core based on the Tensilica Xtensa Diamond Standard 106Micro running at 80 MHz, it has 16 GPIO pins. On the memory front, it has 32 KiB instruction RAM, 32 KiB instruction cache RAM, 80 KiB user-data RAM and 16 KiB ETS system-data RAM. Other features include WEP or WPA/WPA2 authentication.
The advantages of the ESP8266 include its low cost, reliability and easy availability in the market.
Until a while ago, most of its documentation was in Mandarin, although the emergence of the ESP8266 community in recent days has helped translate several details about programming.
Examples of ideal projects include tracking geolocation, building a wireless web server, putting pressure sensors on railway tracks to detect animal presence and set off an alarm (thus avoiding animal deaths on tracks), building smart plugs, humidity and temperature monitoring, and even making a personal assistant of your own (think along the lines of SIRI, Google Assistant, Alexa!).
5. Arduino Uno Rev3
Arduino UNO is an open-source microcontroller board based on the Microchip ATmega328P microcontroller and developed by the tech company Arduino. The board is equipped with sets of digital and analog input/output (I/O) pins that may be interfaced to various expansion boards (shields) and other circuits. It has 14 digital input/output pins, 6 analog inputs, a 16 MHz quartz crystal, a USB connection, a power jack, an ICSP header and a reset button. Since it has everything required for supporting the microcontroller, one can simply connect it to a computer with a USB cable or power it with a AC-to-DC adapter or battery to get started.
Arduino Uno is easy to learn, has low costs, plus a wide array of sensors and several third-party libraries. There’s a large number of (online and free) resources that can be used with this IoT board.
When compared to Raspberry Pi, Arduino Uno needs some effort in order to carry out tasks such as scheduling and database storage. So Arduino is not a great choice if you need more processing power and memory.
Arduino Uno projects
Arduino Uno is perfect for projects such as programming a temperature and humidity sensor, detecting motion with AC/ static electricity, controlling an automated soldering robotic arm and so on.
6. RioRand EP2C5T144 Altera Cyclone II FPGA Mini Development Board
Using the CycloneIIEP2C5T144 chip of ALTERA Company as the core minimum system, the FPGA is easily embedded into the actual system.
Complete with simple logic control, data acquisition, signal processing, mathematical calculations and other functions.
Core board FPGA chip pin all leads, the board can directly into the application board, the 5V power supply board can be introduced through the outlet, and can also be introduced through the row array.
Compact size: 6cm x8cm
On-board EP2C5T144 chip
The EPROM chip adopts EPCS4 with 4Mbit.
On-board 50M ACTIVE CRYSTAL ( on the back)
Power indicator and reset switch
On-board 3 SMD LED
All the IO ports and clock pins lead by the pin, through these interfaces can be extended to any memory and peripheral.
The board uses high quality tantalum capacitors for power supply filtering.
Support the development of NiosII embedded CPU.
7. Particle Boron
A development board that lets one connect a mesh network to over a cellular service, Particle Boron can act as a standalone cellular endpoint or LTE enabled gateway for Particle Mesh networks. Based on the Nordic nRF52840 SoC (System on a Chip), it has a built-in battery charging circuitry. In other words, it’s a lot like the Electron, except with more features such as mesh networking and Bluetooth. It also boasts two SARA cellular modules, NFC, lots of GPIOs, LiPo charging and more. The board can act as a standalone cellular endpoint or an LTE enabled gateway for a Particle Mesh network. The Boron is also configured in the popular Feather footprint and is thus compatible with Feather Shields. Getting a Boron gives one access to the Device Cloud (10 free micro networks up for grabs) along with three free months of device Cloud access using a cellular gateway.
Particle Boron’s big advantage over Particle Electron is that Particle Borons can communicate with each other over WiFi and then pass on the information to a central station which is far away over a cellular network. Moreover, the Boron costs less than the Electron, in addition to offering features such as header pins, LiPo connector, and buttons.
One big disadvantage of moving from Particle Electron to Particle Boron would be the loss of surface mountability.
Particle Boron projects
The Particle Boron is a good choice as a gateway to link an entire group of local endpoints where Wi-Fi is missing or unreliable; it is also ideal for connecting projects to the Particle Device Cloud.
8. ODROID boards
ODROID refers to a series of single-board computers and tablet computers created by the South Korean company, Hardkernel Co, Ltd. The word ODROID is a portmanteau of open + Android, and yet the hardware is not actually open because some parts of the design are retained by the company. Several ODROID systems are capable of running not only Android, but also regular Linux distributions.
The ODROID boards are quite similar to Raspberry Pi, and in many areas, they fare better as well. They typically have a powerful CPU, more RAM and a higher price. For instance, the XU models feature an Exynos SoC even as the C models boast an Amlogic system on a chip (SoC). Both have an ARM CPU and an on chip GPU. While the CPU architectures include ARMv7-A and ARMv8-A, the board memory ranges from 1 GB RAM to 2 GB RAM. Most boards have between three and five mixed USB 2.0 or 3.0 slots, HDMI output, and a 3.5 mm jack. Many of the general-purpose input/output (GPIO) pins offer a lower level output. The latest models have a Gigabit Ethernet (8P8C) port and eMMC module socket. Like with Raspberry Pis, SD cards are used to store the operating system and program memory in the SDHC or MicroSDHC sizes.
ODROID boards are typically way faster than Raspberry Pis, especially upto Raspberry Pi 3. Also, the ODROID-C2 comes with a huge heat sink, another big advantage. Another plus is the fact that the ODROID-C2 board layout is almost exactly the same as that of the Pi B+/2/3. So the largest components (think LAN, USB, HDMI, OTG, GPIO and the screw holes for mounting) are similarly placed allowing someone who has the proper mounts/cases for a Pi to easily work with an ODROID-C2 (in most situations anyway).
In comparison to Raspberry Pi, ODROID boards are more expensive. This makes Raspberry Pis more value for money; for instance the Raspberry Pi 3 has in-built Bluetooth and Wifi. Also, ODROID boards do not enjoy the same quality software and community support that Raspberry Pis do. In fact, it should be acknowledged that the release of Raspberry Pi 4 has only made it tougher for people to opt for ODROID boards.
There is a host of projects one can work on with ODROID boards. Right from building digital photo frames and environmental sensors for game machines to putting together an entertainment system and game kits, ODROID boards are useful for many fun projects.
9. Particle Electron (cellular-enabled IoT board)
Particle is a full-stack IoT device platform with device, connectivity hardware, cloud, and even SIMs for cellular products. This means that when you buy a Particle Electron, you get a cellular module, that allows you to connect with the Internet in over 120 countries, and it is accompanied by a Particle SIM card, a microcontroller, the input and output pins, antenna and USB cable, battery, buttons and LEDs. It has a 120MHz ARM Cortex M3 microcontroller, 1MB flash, 128KB RAM, RGB status LED, 30 mixed-signal GPIO and advanced peripherals plus an open source design.
It complies with any cellular standard and comes with 2G/3G connectivity, SIM card, a low-cost data plan, and some great software for making cellular-connected products. Since the Particle Electron is a pre-configured module, it is ready to connect to the internet straight out of the box.
Cellular connectivity entails license fees and typically, the Particle Electron Iot board is bigger, heavier, more expensive and power consuming than an Iot board that has Bluetooth low energy or Wi-Fi.
Particle Electron projects
The ideal projects include making a GPS + Cellular tracker (to track kids or even valuable objects), monitoring remote sensors and making a high current alarm.
10. Tessel 2 (SBC)
A robust IoT and robotics development platform, Tessel 2 allows one to tap the libraries of the Node.js web development framework so as to make a variety of useful devices within a short span of time. This primary processor of the Tessel 2 runs a very lightweight version of Linux called OpenWRT.
Built to allow the fastest possible production, its USP is its plug and play modules and high-level APIs are its USP. One doesn’t have to set up and maintain a Linux system or track down pinout diagrams to wire up sensors here. Its specs include a 580MHz WiFi router system on chip, 64 MB of DDR2 RAM, 32 MB of flash storage, two high-speed USB 2.0 ports, a micro USB port, a 10/100 Ethernet port (RJ-45 jack) and a 48MHz ARM Cortex M0 microcontroller.
The Tessel 2 looks good but it is not as cost efficient as Raspberry Pi nor does it offer enough additional utility over the latter.
Tessel 2 projects
Tessel 2 has two custom “module” ports from where one can add 10-pin modules like accelerometers, climate sensors, Infrared and more. Each module has several instructions and tutorials available online. From sensing to actuation to connecting with other devices, interacting with the physical world has been made really simple with Tessel 2. Ideal projects are those that include detecting ambient light, detecting humidity and temperature in the environment, playing and recording audio, connecting Tessel anywhere with support for mobile internet on a GSM based network and so on.