You may be overwhelmed by hearing so much about artificial intelligence (AI). However, you will want to read this.
Over the past several years, this groundbreaking technology has worked its way into many areas of day-to-day life. How can AI and automation be applied to inventory management in electronic manufacturing?
The market for electronic components fluctuates constantly, and fast tech cycles mean that parts often become obsolete quickly. Technology plays a large role in managing this inventory effectively and ultimately preventing surplus from becoming electronic waste (e-waste).
“The growth of electronic manufacturing is outpacing the growth of recycling. Proactive solutions with the help of technology should be evaluated to minimise e-waste," says Grant.
This blog explores AI and automation regarding component inventory management with insights from Component Sense Chief Technology Officer, Grant Rutherford.
AI in inventory management
IMB, in collaboration with Oxford Economics, surveyed 2,200 executives in 2023 and found that 89% said that crucial investments in automation will include generative AI capabilities. AI generally refers to machine learning, large language models (LLMs), or predictive analytics.
AI is currently being utilised to drive efficiencies for various tasks in electronic manufacturing, including general data analysis, research, and copywriting. Forecasting AI technology remains the most promising use case, however, currently, it is in its infancy and is not yet fully capable.
The potential of AI forecasting
By utilising historical sales data, a knowledge base of general customer behaviours, and seasonal trends, AI can look for patterns and attempt to predict future stock demand. It can also be utilised for tracking current stock levels, recognising sales patterns, and identifying alternative components quickly, all of which aid the procurement process.
If AI is integrated well, and as AI technology continues to evolve and become more contextually aware and responsive, it may be able to link to live market data and social trends. The potential to forecast in real-time using live data is an exciting prospect.
As well asidentifying alternatives for recently obsolete parts, these real-time suggestions could include ways to repurpose stock across multiple company sites. Ultimately, more accurate forecasting prevents stock overages that contribute to electronic waste. However, it must be said that the link between AI and sustainability is complex, given the high energy and water usage required for data centres.
"I fully encourage companies embracing LLMs and AI in general to improve efficiencies and reduce waste wherever possible. Unfortunately, due to the nature of forecasting, AI has limited use in preventing excess as it stands,” explains Grant.
Challenges regarding AI forecasting
Forecasting is inherently challenging because the future is unpredictable. Variables such as obsolescence, sudden tech advancements, geopolitical unrest, and environmental disruptions often arise without warning.
"If a company expects to sell 30,000 units but only sells 10,000, AI cannot accurately foresee this outcome. The same challenges make forecasting difficult for humans. While AI forecasting tools do exist, their effectiveness remains to be seen. As is the case when using AI in general, poor data quality can also lead to inaccurate results. Overconfidence in AI for forecasting may lead to excess and obsolete stock accumulation and spending more than is required. Even the best AI models can produce inaccurate or misleading results (hallucinations),” explains Grant.
As the popularity of zero-trust supply chains increases, security concerns regarding sharing sensitive inventory data with AI should also not be overlooked.
Automation in inventory management
Whether utilising new software or machine robotics, automation aims to carry out repetitive or routine tasks based on a set of rules. Ultimately, automation aims to reduce errors and complete tasks more quickly.
Current everyday use cases for automation include tracing inventory and reporting, reorder point automation, invoicing, and physical robotics for picking and packing stock. However, these examples are just the tip of the iceberg.
A McKinsey & Company report explains that some expert sources expect warehouse automation to grow by more than 10% annually until 2030. Automation allows staff to focus on more strategic, creative, and problem-solving-based work.
The potential of automation for inventory management
As warehouse robotics evolve and become more complex and capable, their ability to speed up fulfilment, restock, and pack increases. New cutting-edge robotics can enhance existing systems and drive further efficiencies.
IoT (Internet of Things) sensors are already being used in many electronic manufacturing warehouses to provide real-time visibility across multiple sites and track inventory. More IoT-connected devices could further improve visibility, offer clearer operational insights, and enable better-informed inventory decisions.
"Automation can enhance operational efficiency by reducing manual data entry and processing. AI-integrated automation is also an exciting prospect, with AI potentially being able to trigger and set up automation in real-time," explains Grant.
Challenges regarding automation in inventory management
As with any tool, automation requires human oversight to handle issues or unexpected scenarios. Automating a process or system that was flawed to begin with simply speeds up an inefficient system.
Like AI, implementing automation often requires a significant upfront investment, mainly due to the range of MRP systems. However, in many cases, this initial investment will generally pay for itself and more in the long run.
Utilise technology in your inventory management strategy
Both AI and automation play valuable roles in inventory management and forecasting, with their capabilities set to grow as technology advances. However, companies should be realistic about what is achievable today regarding AI.
"Many companies are still simply scrapping excess stock regardless of whether AI is being used for forecasting, so a better solution that addresses both sustainability and efficiency challenges is needed," says Grant.
Component Sense’s InPlant™ excess and obsolete component inventory redistribution system utilises automation and bespoke software to actively identify excess stock at the earliest possible stage. This approach allows for quick redistribution and maximised financial returns for you.
"Component Sense can customise its services to fit customer and supplier needs. The IT team has developed bespoke software to handle nearly every aspect of operations. InPlant™ can be moulded to fit your needs. Component Sense brings together custom technology and expert human processes to deliver scalable, ethical E&O inventory solutions," outlines Grant.