printer friendly versionThree reasons why AI, ML add value for SMMEs only if the basics are in place
26 July 2023
AI & ML
There is much chatter around artificial intelligence (AI) and the subfield of machine learning (ML), which can be confusing for SMME owners who may believe that they need to climb on the bandwagon. That’s why it’s time for a reality check.
When SAP first introduced the concept of the intelligent enterprise, it was defined as: “An intelligent, sustainable enterprise is one that consistently applies advanced technologies and best practices within agile, integrated business processes.”
ERP systems play a crucial role in enabling the intelligent enterprise. An intelligent enterprise is one that leverages data, analytics, and digital technologies to optimise its operations, but does this mean that AI is needed in the business?
ERP systems are designed to help SMMEs manage their operations and processes more efficiently by integrating various departments, automating routine tasks, and providing real-time data insights. While AI and ML can enhance these capabilities by analysing large volumes of data and predicting outcomes, their implementation can also be complex and expensive.
Advanced technologies like AI, ML and the Internet of Things (IoT) are powerful tools that can be used to solve a wide range of problems. “But to effectively leverage these technologies, it is critical to first have a solid ERP foundation in place to integrate data, infrastructure, and business processes. Without the basics in place, any business challenges that the organisation is trying to address will not be resolved.
Before SMMEs think of looking at AI, they need to build the basics, which include centralised data, automated tasks, technology integration and real-time insights that enable SMMEs to grow and be profitable. Here are three reasons why advanced technologies are useful and appropriate only when the basics are in place:
1. Quality data is essential.
AI and ML algorithms rely on large amounts of high-quality data to learn and make accurate predictions. If the data is incomplete, inconsistent, or inaccurate, the results of the AI or ML model will be similarly flawed. That’s why it’s crucial to have a robust data collection, management, and quality assurance process in place to ensure that the data is clean, reliable, and suitable for use in machine learning.
2. Infrastructure and computational resources.
AI and ML require a significant amount of computational power and infrastructure to run efficiently. Without proper infrastructure, including hardware and software, the algorithms will not be able to run quickly or accurately. Moreover, this can result in increased operational costs and decreased accuracy in decision making.
3. Business processes.
Sophisticated technologies must be integrated into existing business processes to be truly effective. Organisations must have a clear understanding of their business goals, the problems they are trying to solve, and the metrics they use to measure success. Without these foundational elements in place, AI and ML may be unable to provide meaningful insights or actionable recommendations.
AI and ML are terms that refer to the use of technology to model human intelligence. They are the current buzzwords, just as the cloud once was. That’s not to suggest that they are not powerful technologies, but simply to underline that they will not solve business issues if they are not deployed on top of an existing infrastructure that works. Much like ChatGPT, they will not provide all the answers people are looking for if they are not applied correctly, on top of operations that are running optimally, and in harmony with a well-designed ERP system.
There’s no doubt that businesses across all sectors will continue to embrace AI and ML technology over the coming years, transforming their core processes and business models to take advantage of machine learning for enhanced operations and greater cost efficiencies.
To make the best use of this technology, we suggest beginning by spending time on developing a use case that defines and articulates the problems or challenges that the business would like AI to solve, and then to ensure the processes and systems already in place are capable of capturing and tracking the data needed to derive real value from the technology.
Without ensuring this, the organisation will gain bragging rights with no value add. If the company does not have the processes and systems to drive efficiencies it will be unable to leverage the promise of the technology to grow the business and that means the project has failed.
For more information visit www.seidorafrica.com
Further reading:
Give your edge AI model a performance boost
AI & ML
Join this webinar from STMicroelectronics to learn how to create an edge AI application easily on an STM32 MCU using the NVIDIA TAO toolkit.
Read more…
Game-changing graphics innovations at the Edge
Rugged Interconnect Technologies
AI & ML
With an outstanding price-to-performance ratio in its class, ADLINK’s MXM-AXe offers competitive pricing that rivals the renowned NVIDIA T1000.
Read more…
Portable GPU for AI applications
Editor’s Choice AI & ML
The Pocket AI module consists of an Nvidia RTX A500 GPU with 4 GB GDDR6 RAM onboard, 2048 CUDA cores, 64 Tensor cores, and 16 RT cores embedded.
Read more…
AI-based visual inspection system
Avnet Silica
AI & ML
The Defect Visual Inspection solution is a combination of a compact system-on-module and a software library optimised to run at the edge, making it ideal for applications in industrial, medical, food, electronics, semiconductor, and packaging sectors.
Read more…
Arm Cortex-M7 MCU product family
NuVision Electronics
AI & ML
GigaDevice Semiconductor Inc. introduces its first Arm Cortex-M7 core microcontroller product family, the GD32H737/757/759 ultra-high performance MCU series.
Read more…
Sub-GHz SoC with built-in AI/ML accelerator
Altron Arrow
AI & ML
A dual-band SoC, the FG28 includes radios for sub-Gigahertz (GHz) and 2,4 GHz Bluetooth LE, and a built-in AI/ML accelerator for machine learning inference.
Read more…
First South African QBronze workshop
AI & ML
The QBronze109 workshop is titled ‘Quantum Computing and Programming’, and participants will learn the basics of quantum computing and how to write simple quantum programs.
Read more…
Giving you the smart edge
RS South Africa
AI & ML
The i.MX 93 system-on-chip architecture integrates one or two Arm Cortex-A55 cores, one Arm Cortex-M33 core, and an Arm Ethos-U65 Neural Processing Unit.
Read more…
What is ML? – Part 3: Hardware conversion of convolutional neural networks
Altron Arrow
Editor’s Choice AI & ML
In this series, the CIFAR network, with which it is possible to classify objects such as cats, houses, or bicycles in images, is discussed. Part 3 explains the hardware conversion of a CNN and the benefits of using an AI microcontroller with CNN accelerator.
Read more…
AI-Inferencing small form factor computer
AI & ML
Designed for video and AI applications at the edge, the FALC product family offers rugged and industrial variants, is suited for any environment, and is ideal for a wide range of applications and industries.
Read more…
