Raspberry Pi Vending Machine

According to tomshardware.com, iSnowfall (Reddit ID), in the final year’s project for students with special needs, has produced a 1.8-meter-high and .5-meter-wide vending machine powered by Raspberry Pi. It uses RFID cards for transactions. The RFID UID number is stored on a database on a Google sheet via a Python script. Once the Raspberry Pi confirms the points received, it drives the motor to deliver the selected products. See Figure 9-2.

The vending machine uses a PCB integrated with Raspberry Pi Compute Module and other essential component boards and accessories. You can also attach a Hat with all the functions you need on the Raspberry Pi.

Upstate Networks Inc. (UNI), a hardware and software manufacturer from New York, used the Raspberry Pi to develop vending machine peripherals. Since 1994, UNI has designed adapters that bridge vending machines and standardized computers. On the UNI website, you will find a product line called PLUM, which is an unattended sales system integrated with a Raspberry Pi-embedded controller. See Figure 9-3.

Interactive Touchscreen Directory for a Smart City

There are more than 8,000 locals and tourists visiting Coogee Beach in Australia. There are awesome bars, restaurants, promenades, and parks everywhere. In 2018, the Randwick City Council adopted a smart city strategy seeking to improve livability, sustainability, and economic prosperity through digitalization and IoT. This strategy established a roadmap to guide and accelerate the smart city action in Randwick. The Randwick Smart Beach Project faces the challenge of creating a better beach experience: applying public safety, traffic, parking management, facility maintenance, and all open data on the beach with a public dashboard. In this case, Coogee Beach adopted Ubidots to create dashboards to view sensor results and control devices in real time. See Figure 9-4.

To develop a custom dashboard, you can simply populate the dashboard with prebuilt graphs, charts, plug-ins, indicators, maps, and control units, or use HTML Canvas and custom code.

Figure 9-5 shows that Peclet developed a digital display and integrated all these tools in Ubidots. The display shows the data from all IoT sensors placed around Coogee Beach. The sensor data is transmitted through LoRa used with the LoRaWAN web server through “The Things Network”.

Raspberry Pi as a Retail Product Display

Digitec is an electronics retailer in Switzerland. Among other things, they sell Raspberry Pis and related accessories, including their official 7-inch touch display. Many customers might have noticed that they haven’t had the touch display in stock recently, and there’s an interesting reason for that. See Figure 9-6.

The retailer wanted to replace their tablet-based digital shelf labels with something more robust, so they turned to Raspberry Pi 2 with the 7-inch touchscreen. Each store has 105 screens, which means that the engineering team for Digitec Galaxus assembled 840 custom-printed Raspberry Pi-powered shelf labels to replace the existing paper-based product labels. See Figure 9-7.

The screens enable their customers to view up-to-date product information, prices, and customer opinions based on community ratings as they see the product up close. To pull this off, the engineering team used Raspbian Jessie Lite and installed the Chromium browser. It enabled them to create a custom HTML page that loads JavaScript to download the most up-to-date information using a JavaScript Ajax call. When a keyboard is connected, it can set parameters for the screen, which are stored as cookies in Chromium. The engineering team also introduced unnecessary redundancies into the design. For example, the boot script will put Chromium in a loop to ensure that it will eventually be relaunched if it crashes. It is also capable of handling sudden loss of power and network connectivity issues.

A Multimedia Toolkit for Museums, Visitor Centers, and More Running on the Raspberry Pi

Pi Presents is a toolkit for producing interactive multimedia applications for museums, visitor centers, and more.

There are a number of digital signage solutions for the Raspberry Pi, which are generally browser based, limited to slideshows, non-interactive, and driven from a central server, thus enabling the content to be modified frequently. See Figure 9-8.

Pi Presents is different. It is standalone, multimedia, highly interactive, and diverse in its set of control paradigms—slideshows, cursor controlled menus, radio buttons, and hyperlinked shows. It’s able to interface with users or machines over several types of interfaces. It is aimed primarily at curated applications in museums, science centers, and visitor centers.

Being so flexible, Pi Presents needs to be configured for your application. This is achieved using a simple-to-use graphical editor; it needs no Python programming. There are numerous tutorial examples and a comprehensive manual.

Pendulum Wave Machine: A Museum Exhibit

The University of Florida Physics Department has a lobby exhibition which is very reminiscent of the Discovery Gallery in the London Science Museum. One of their latest exhibits is a Pendulum Wave Machine. They used the GPIO and animation facilities of Pi Presents to provide tightly coupled user interaction with the machine. See Figure 9-9.

Raspberry Pi 4 Cluster for Hosting a Raspberry Pi 4 Website

An 18-board Raspberry Pi 4 cluster was used to host much of the official raspberrypi.org website on its busiest day. The recently released 4GB Pi 4 costs $55, a fraction of the cost of traditional server hardware, but is also less powerful and offers less memory than entry-level servers. Mythic Beasts, the ISP that hosts the Raspberry Pi website, configured the cluster to serve most of the website and found it handled record demand on the day of Pi 4’s release without issue. See Figure 9-10.

The Raspberry Pi foundation ran the website on this Raspberry Pi 4 cluster for over a month before reverting back to the usual virtual server-based environment.

Use Virtual Network Computing (VNC) to Remotely Control Everything You Want to Execute on the Raspberry Pi

Sometimes it is not convenient to work directly on the Raspberry Pi or access it using SSH with Visual Studio Code or OpenSSH client. If you want to edit the code from another device and need to access the graphical user interface of Raspberry Pi, we suggest trying RealVNC for accessing a remote desktop. VNC is a graphical desktop remote-sharing system that allows you to remotely control the desktop interface of a computer (running the VNC server on the Raspberry PI) from another computer or mobile device (running the VNC viewer on your laptop). See Figure 9-11.

You will be able to see and control the Raspberry Pi desktop in a window on your computer.

VNC Connect from RealVNC is already included in the Raspberry Pi OS Desktop version. Download the VNC Viewer from their website first to the device you want to use to remotely control your Raspberry Pi, such as your Mac, Windows machine, or Linux P. See Figure 9-12. Then you can use the following simple tutorial to enable VNC and use it on your Pi.

Make your PC join the same private local network as your Raspberry Pi (for example, an Ethernet or WiFi network).

On your Raspberry Pi, discover the private IP address by double-clicking the VNC Server icon or running ifconfig from the terminal. See Figure 9-13.

On your PC, run VNC Viewer and enter the IP address that you just found for the Raspberry Pi in the search bar, as Figure 9-14 shows. After successful connection, you will see the Raspberry Pi desktop on your PC. You are now ready to use your development board remotely with access to the graphical user interface!

IoT Integration Platform for the People-Counting Project

Chapter 2 introduced computer vision and deep learning, which were used to develop a people-counting application . However, in order to include an end-user friendly dashboard that can deliver all the data you want to analyze for the decision makers’ benefit, you need to merge the physical camera results with the digital analysis into one user interface.

Thousands of IoT entrepreneurs, startups, and system integrators are using Ubidots to rapidly launch and scale Internet of Things (IoT) businesses without having to write code. An IoT integration platform will save you a lot of time and money when creating and testing a visualized operation dashboard that connects multiple hardware devices. The complexity of deploying an end-to-end complete IoT solution goes far beyond the application prototype, which requires a variety of IoT technologies such as hardware compatibility, communication, software, and UI interfaces. If you have confirmed that the best solution is using hardware, a connection, and the cloud—such as using a Raspberry Pi connected to the cloud via WiFi—you still need to develop an application to provide data to your customers and control all the devices from the visualized dashboard.

For IoT applications, the last layer involves backend and frontend components, and it usually takes thousands of engineering hours to deliver a stable solution with a flexible UI interface. In order to quickly assemble and launch IoT applications, you can choose an IoT integration platform like Ubidots, which has both hardware and software options and efficient integration tools. See Figure 9-15.

Real-world data is critical to digital transformation. How to collect and process data remains a challenge for system integrators, device providers, business-end users, and many stakeholders. For retailers, it is still difficult to use automation and digitalization to drive their businesses from ideas into practice. End-to-end IoT solution platforms, such as Seeed, help retailers and end users accelerate and scale their IoT solutions from hardware end to user end, offering real-world data collection from nodes, edge devices connection with multiple communication choices, data operation at the cloud, and SaaS and PaaS to achieve digital transformations. See Figure 9-16.

Choosing an IoT integration or solution platform can speed up retail business automation transformation. Its low cost and power consumption usually are essential to business retailer facilities as vending machines and customer service directional devices.

Microphone Options for Speech Recognition

Different microphone options in Raspberry Pi have many factors to take into account, including size, cost of development, and maintenance cost. This makes voice recognition design more challenging. While any microphones can work for your proof of concept (PoC), they will likely produce poor-quality results with your Raspberry Pi voice recognition software.

To overcome these challenges, Seeed Studio provides the reSpeaker USB microphone array, as shown in Figure 9-17. It offers a better voice pick up, which is essential when you place the drive-through machine outside the restaurant. The reSpeaker USB microphone array supports up to 5m far-field voice pickup and 360° pickup mode, and it implements the following acoustic algorithms: DOA (Direction of Arrival), AEC (Automatic Echo Cancellation), AGC (Automatic Gain Control), and NS (Noise Suppression).

Cluster Server Carrier Board

As mentioned in Chapter 8, Raspberry Pi 4 is capable of serving a website to millions of visitors, if several boards are combined together in a cluster. While it is feasible to build a cluster yourself, just like we did for the prototype project, by connecting Raspberry Pi 4s with Ethernet cables, Turing Pi offers a better alternative, called a “cluster on board.” There are two models of Turing Pi currently on the market: Turing Pi 1 supports Compute Modules 1, 3, and 3+, and the upcoming Turing Pi 2 also supports Compute Modules 4. See Figure 9-18.

Turing Pi is a compact ARM cluster kit that provides a secure and scalable compute in the edge. It is designed to make edge computing easier for developers. Turing Pi cluster architecture allows you to migrate and sync web apps with minimal friction.

Turing Pi includes dedicated I/O connections for the first slot, so you can manage the entire cluster through a Pi in one slot, or you can use another computer to manage the entire cluster externally. You only need to plug in the power supply, keyboard, and mouse (or the power supply and network cable) and you are ready to use it! See Figure 9-19.