Computer/Embedded Technology

AI platform and digital signage for vending machines

28 September 2022 Computer/Embedded Technology

The adoption of IoT and smart technologies in industrial and manufacturing settings is well established but the trend toward digitalisation is expected to increase even more in the coming years. Interaction with intelligent machines at work, at restaurants or while walking on the streets will soon become commonplace. Intelligent machines are going to change the way we consume and experience the world. way we consume and experience the world.

The evolution of the machines and displays we interact with daily, often without even paying attention to them, will be at the heart of this transformation. Not only smartphones and computers, but automatic checkouts at the supermarket, beverage dispensers in the office, and interfaces to manage machinery at work will need to evolve. The interaction between increasingly intelligent machines and humans will become more immediate and natural.

The latest developments in digital signage and vending machines are a case in point of the digital revolution underway. They will revolutionise the way businesses handle customer relations and the customer experience in general.

AI for digital signage

Digital displays capture the attention of potential customers and, due to their visual impact, have better results in influencing behaviour, increasing time spent in a store, and encouraging customer intentions to revisit in the future

AI-based digital signage implemented with cameras, microphones, sensors and integrated into a digital network could quickly examine the characteristics of the person in front of it and suggest a product to buy, or, better yet, it could display a message that changes with the gestures and behaviours of the person

These kinds of digital signs can bridge the gap between the online and offline customer experience by assisting in the purchase process and making recommendations about the products. For companies, it means being able to make the customer experience more pleasant and to have direct information about users.

The right technology for smart machines

Building smart appliances needs the implementation of suitable software and the HMI such as advanced AI, touch displays and platforms to manage and process data. This is where SECO technology stands out for its products for data science, AI and edge computing applications. Thanks to its expertise in building IoT systems and sophisticated sensors, SECO manages to develop highly innovative solutions in both hardware and software for digital signage.

SECO products are integrated with predictive maintenance solutions which signal any anomalies and guarantee rapid intervention in the event of damaged components. All of this takes place together with ensuring both high graphics performance and low power consumption.

SECO’s platform for leveraging smart device data, CLEA, is a flexible and powerful tool and great for helping digital signage companies digitally transform their devices. Combined with SECO’s data collection and aggregation platform, Astarte, CLEA allows you to turn collected data into insights or to use it to create richer, more interactive, personalised interactions.

Technical application

A project demonstrating the power of the combination was developed. The project was for a vending machine that handles the interface, data collection and beverage dispensing and the cloud application hosted on the CLEA platform.

The application has a simple interface that shows the real-time feed from the camera and the suggested products. The machine uses a CNN (convolutional neural network) to detect the face and estimate the gender, age, and emotions of the user. Once a beverage is selected, the machine starts dispensing and uploads the data to the Astarte cloud platform which is part of the CLEA ecosystem.

CLEA then allows for an easy retrieval of all recorded information, showing tables, charts and other statistics or aggregated data through a cloud application.

The language used to develop the vending machine application is Python 3 using PyQT as the user interface framework. The face detection and the gender, age and emotion estimations are performed with a few ML (machine learning) models in ONNX format, accelerated through an OpenVino runtime environment to exploit the Intel hardware available onboard the SECO device.

The use case presented in this article is just the tip of the iceberg when it comes to deploying AI in typical signage applications where interaction needs to be handled on a per user basis.


Share this article:
Share via emailShare via LinkedInPrint this page

Further reading:

Avnet Silica expands its embedded vision solutions portfolio
Avnet Silica Opto-Electronics
Collaboration with camera module manufacturer Allied Vision includes a complete set of mainline drivers in the Linux kernel.

Touchscreen controller family gets additional security features
Avnet Silica Opto-Electronics
Microchip’s ATMXT2952TD 2.0 family of touch controllers now offer cryptographic authentication and data encryption.

Easily incorporate embedded security
Avnet Silica DSP, Micros & Memory
Microchip Technology has announced the new family of PIC32CK 32-bit microcontrollers that incorporate an integrated Hardware Security Module (HSM) subsystem.

Adaptive computing for big data sets
NuVision Electronics Computer/Embedded Technology
The new AMD Alveo V80 compute accelerator, powered by the AMD Versal HBM adaptive SoC, is built for memory-intensive workloads in HPC, data analytics, network security, storage acceleration, and financial technology.

MCUs offering more design freedom
Avnet Silica DSP, Micros & Memory
The STM32H7R and STM32H7S are the fastest STM32 MCUs at 600 MHz, and the most cost-effective STM32H7s, thanks to their smaller flash memory.

NeoCortec wireless mesh networking
CST Electronics Computer/Embedded Technology
The protocol is especially designed to allow real-time routing of data over long distances and multiple hops, while using very little energy.

GasVisor wins Advanced Electronics Challenge
Avnet Silica News
Avnet Silica announced GasVisor as the winner of the Advanced Electronics Challenge, an initiative which champions start-ups and young, innovative companies with hardware-based product ideas.

Embedded software development
Computer/Embedded Technology
The reliance on C is being reduced, with Python the language of choice for embedded applications in the fields of IoT and AI.

Microsoft Windows IoT on ARM
Altron Arrow Computer/Embedded Technology
This expansion means that the Windows IoT ecosystem can now harness the power of ARM processors, known for their energy efficiency and versatility.

8-bit MCU with I3C support
Avnet Silica DSP, Micros & Memory
The PIC18-Q20 8-bit microcontrollers from Microchip easily interface with devices operating in multiple voltage domains, and the built-in I3C interface supports higher-speed and lower-power data transfers than I2C.