Oct. 27, 2023
Using AI to Improve Your Sales & Operation Planning
Artificial Intelligence (AI) is a technology that can be intimidating for small and mid-sized manufacturers. Many assume it is out of reach in terms of costs, implementation and required knowledge. But AI doesn’t have to be expensive or complicated. There are many AI tools designed to be user-friendly to help you harness their power.
“Utilizing data to drive business decisions starts with collection. Whether you want to use existing data from your business system, real-time production information from the shop floor, or industry metrics to make decisions, it is critical to make sure the data source is organized, true, and scaled appropriately. Well-organized data will produce better outcomes regardless of the analyzing party: a manufacturer, data scientist, or AI. The best practices of governing incoming data, connecting IT-OT systems, and organizing the stored information will continue to lead to better results.”
Cory Larson, CMMC-RP, WMEP IT-OT Lead
AI can be generative or predictive. Generative AI, such as ChatGPT, creates something new (like writing a paper or a custom email). If you use a “custom DJ” to make a playlist in a digital music service, you’re in essence using Generative AI. Predictive AI identifies patterns, variables and trends and makes predictions and/or takes actions. Some of the human decision-making is then removed, and the AI-processor takes over. Predictive AI can be a bit like having a smart assistant for your operation.
Predictive AI is the primary focus of this article, although predictive and generative AI can work together. Predictive maintenance is Predictive AI in action. Your veteran CNC machinist may know how far she or he can push that vibration in the machine before doing maintenance. AI tells a brand new machinist when maintenance should be performed to prevent downtime.
A great starting point for use of AI is to improve your Sales & Operation Planning (S&OP) results. Many companies struggle with S&OP because it relies on cross-department collaboration – a skill that many organizations find difficult. Sales and operations often have conflicting priorities; how many SMMs use their sales forecasts as their actual production plan? That makes it difficult to develop and execute an appropriate plan.
AI can analyze historical data, market trends, and many other variables to develop a custom, predictive plan for your operation. It’s custom in that it uses trends, datasets and past outcomes that are unique to your actual supply chain performance. AI is no longer a concept. The time has arrived for manufacturers to start seriously looking at AI and the right partner can help you get started.
Think About AI as the Intersection of Institutional Knowledge and Data Insights
Many manufacturers are surprised to learn how AI can help capture institutional knowledge, almost like a digital memory of your experienced staff. It can learn from their actions and resulting outcomes. So even when those experienced staff members retire or leave, their knowledge stays with your company, helping new employees make better decisions.
Think about the institutional knowledge in a small operation, say 12 to 13 people, and how that differs from a manufacturer with about 100 employees. The office manager in the small operation seems to know nuances about everything; the supplier who struggled with on-time delivery in Q1 because of weather, the supplier whose delayed response always triggers an expedited order; or the FedEx driver who always arrives late on Thursdays and Fridays. These same conditions may exist in a 100-person company, but that information is dispersed more widely, which can make it harder to access and interpret.
Predictive AI can help you manage these types of exceptions with suppliers. Imagine AI as a watchful eye that can spot issues with your suppliers before they become big problems. It can detect patterns that indicate a supplier might be late or is having quality issues, giving you time to make alternative plans.
Let’s say you are a job shop that handles 1,000 orders a week. That’s about 52,000 actions a year, which would be a lot of data to collect and analyze for any human, even with well-constructed dashboards. AI will process all of that and provide insights, including variations and trends. It’s a great example of where AI can help you build on your foundation of data to replace some human decision making for S&OP.
The History of Decision Aides Such as ERPs and MRPs
Manufacturers have often brought together operations, production, purchasing, and even engineering to compare different viewpoints and data sets. This approach provided a forum for each respective organization to talk through their challenges. Some companies include sales in these discussions. The result is usually an improved but far from optimal S&OP, but at a high resource cost.
The industry got a little better at this process with decision aids built around tools such as Enterprise Resource Planning (ERP) or Material Requirements Planning (MRP) software to align supply and demand. These systems processed more information in a shorter time. You had a more powerful tool, but it still required people to analyze the information and make decisions, and the data outputs are based on static variables which many manufacturers seldom update (e.g., product and/or shipping lead times).
These tools further evolved to include dashboards and plug-ins to help manage and identify exceptions and help humans make better decisions. The MRP generated planning purchase orders, suggesting what you should buy. But the information overload component of these tools have contributed to manufacturers simply ignoring some of the analytics or lacking the bandwidth to process what the data is telling them – and thereby failing to reduce their dependence on institutional knowledge or dramatically improve their planning process overall.
Many manufacturers have never fully leveraged the full potential of their ERPs. But technology has continued to evolve, and depending upon your AI solution, it could truly offer the “next level” results that you may have struggled to achieve with previous toolsets.
AI Boosts Resiliency With Better Forecasting and Inventory Management
One of the clear benefits for manufacturers is how AI can help you manage risk in your supply chain. Data about your suppliers’ performance may tell you using supplier A at this particular time, and with this particular order, is a better choice than supplier B. Maybe it’s a large order, which supplier A has struggled with in the past. Or it’s a rush order that supplier B has delivered without fail. Or it might tell you the optimal time to buy a certain material or part. While these are simplified scenarios, they underscore the day-to-day types of decisions for which AI tools could be appropriate.
Another benefit of AI solutions is that the historical data can be visible throughout your organization. In some ways, it replicates your institutional knowledge and shares it widely throughout your organization.
AI can use your historical data to help predict:
- Which SKUs are higher margin
- Performance of your proposed product mix, such as sizes, colors, flavors or varieties
- New product performance based on your previous product launches
- Which aspects of your marketing and sales might be more effective with a new product
- What supply plan is most likely to be successful
AI can help you grow your business and improve your supply chain resilience through better demand forecasting and inventory and supplier management. AI can help you reduce time-consuming exceptions, which frees people up to do more valuable tasks; this can be particularly impactful given the current workforce shortages that many companies are experiencing. Ultimately, AI can also help you improve your fill rate, an important metric embraced by many SMMs.
How to Get Started and Carve Out Key Information
The first step on your AI journey is simply to get educated. Pull a small team together to jointly research what tools are available and how they can help you. Put off any buying recommendation for now. Lean on your sales reps. Create a project and calendar around this education; meet regularly as a team to determine which platforms to research, who to contact, what you have learned.
As you evaluate platforms, pay close attention to the overall usability and data requirements as these factors will often make or break any digital transformation. Your team has to be excited about the platform’s potential and understand how the AI tool could reduce the burden on your staff while improving your overall metrics. You will need to be able to answer these questions:
- How will your staff use the tool?
- What datasets are required to enable it?
- Who maintains these datasets?
Your Local MEP Center Can Help You Get Started on Your AI Journey
AI tools are not out of reach. In fact, you may even have access to AI in your current tech stack. Talk to your customer relationship management (CRM) provider to see if AI tools are offered or to find out what is available that may integrate with your system. There may also be AI functionality available from your current ERP or MRP systems provider.
Technology costs are coming down, in many cases, and technology can help you address a variety of challenges. Look ahead a couple of years: if you’re like most manufacturers, you likely have an aging and underused technology infrastructure, and you will be losing key personnel. Implementing AI is one way to leverage technology to replace that knowledge and enhance your competitive abilities. Your local MEP Center, WMEP Manufacturing Solutions is the MEP center in South East Wisconsin, can help you find the right AI partners and solutions to help you address these challenges.
About the author
Dean Hettenbach is a Senior Extension Professional at Georgia Manufacturing Extension Partnership at Georgia Tech, part of the MEP National Network. In this role, he helps clients turn technology, supply chain, and manufacturing strategies into deliverables that impact how clients operate, scale, and compete. He has expertise in Manufacturing Technology, Supply Chain, Medical Device Manufacturing, Strategy and Leadership Development.