Barcodes have been around since the mid-‘70s, when they were first introduced into the retail environment. Today, it’s hard to imagine a supply chain or stock control system without them. But newer technology is fast threatening the continued existence of the faithful numerical striped mark.
Radio-Frequency Identifications or RFIDs are small chip and antenna identifiers that attach to a product or object. Like barcodes, RFIDs too need to be scanned to retrieve their information. The difference is they don’t need to be positioned precisely relative to the scanner, they can work up to 20 feet away! Imagine placing your bag of groceries near a scanner to instantly total your purchases.
This technology is not without flaws however. The cost is higher than barcodes, and it’s also prone to reader collision (two or more readers overlapping).
A Quick Response Code or QR Code is a matrix barcode (two-dimensional barcode) first designed for the automotive industry in Japan. QR Codes have gained popularity in recent years due to their fast readability and greater storage capacity compared to their one-dimensional cousins.
However, what has really boosted their fame is the proliferation of smartphones. Instead of using traditional scanners, employees can now use the smartphones they already have in their pockets, to track inventory items.
With SYSPRO’s mobile application SYSPRO Espresso, which has access to a devices’ native functionality such as the camera, makes usage of QR codes possible. The Innovation Lab at SYSPRO has also experimented with implementing this in core Espresso to make it even more accessible to users. SYSPRO Espresso also has a highly configurable barcoding framework that gives customers the flexibility to store how their barcode data should be stored from custom form fields, other master columns or even third party tables.
Already in wide use on many social media platforms (Snapchat, Facebook etc.) for facial recognition and search functionality, IR is fast revolutionizing the way we live and work. Take DuLight for example, a program for the visually impaired that recognizes what is in front of the user and then describes it back to them.
In the business world, IR is poised to equally transform the way we do things. Cutting Edge ERP solutions (like SYSPRO) are already accommodating this technology to assist with things like stock control and asset management. Imagine simply viewing an item with your mobile device camera to immediately know product specs, prices, stock levels.
Computer Vision and Image Recognition at SYSPRO
In the past, using Image Recognition and computer vision in software required bespoke Neural Networks or other complex machine learning algorithms to first be implemented. These would require an army of data scientists with PHDs to develop and even then required an immense training dataset of diverse images to make it even vaguely accurate. Now thanks to advances in machine learning and the release of sophisticated open source pre-trained Neural Networks such as Microsoft’s RESNET, computer vision is made far more accessible.
These advances have been central to our approach to Machine Learning and Artificial Intelligence at SYSPRO where, to quote Google Scholar’s (Google’s search indexer for academic journals) homepage, we “Stand on the shoulders of giants”. In this regard, we build off existing Neural Networks as well as sophisticated prebuilt machine learning APIs in the cloud such as Microsoft’s Azure Cognitive Services, and expose these services in an accessible fashion to SYSPRO users. In the Innovation Lab, we have also developed a SYSPRO Business Object that allows for the execution of both R and Python scripts (the most prominent programming languages in artificial intelligence) which makes anything truly possible. With this approach, we hope to provide the power of artificial intelligence to SYSPRO users of all skill levels from beginner to experienced data scientist.