How AI can help manufacturers deal with disruption

The COVID-19 pandemic, and resulting supply chain disruption, have underlined the inability of manufacturers and supply chains to move with agility and to match supply with increasing and changing demand. And, crucially, the inability to make the key decisions within their operations in order to win.

This is where artificial intelligence (AI) can really help – understanding demand signals, aligning production, pricing dynamically and delivering the right product to the right place at the right time. This is how businesses, and manufacturers, can win in today’s increasingly-competitive world.

BDO found that 70% of consumers would be more willing to buy from a brand they knew used AI to manage their supply chain. Here are three in-depth areas to explore, highlighting where innovation and technology can help to drive manufacturing efficiencies.

Understanding demand

Historically

Demand models are traditionally centered around basic trend information and seasonality inputs. This means that consumer demand is usually gauged using internal historical data for products, as well as fixed seasonality, rather than adapting to trends easily.

In many businesses, demand forecasting relies on skilled individuals using their business knowledge to give their best estimates on top of basic, backwards looking data which can often be out of date.

It’s also true that data sits within siloed areas of the business, and is either not used, or used in isolation, rather than being combined and considered as part of a wider data set. This is usually because people and systems are not aware of all of the relevant information held in the business, and because it’s difficult for a human to consider all data in their forecast and decision-making processes.

As a result of this, forecasts are not as accurate as they potentially could be, and the downstream impacts can be significant – particularly on inventory and servicing demand.

Using AI

AI-powered demand forecasting can significantly help businesses to drive action, specifically with a view to making key decisions. Peak has the ability to ingest and combine data quickly and timely, to allow stakeholders to base their decisions on the most recent and available data within the business.

This is done by using data typically held in siloed areas of the organization and leveraging business knowledge, to join it together and create a DI-ready data set – from which a more accurate forecast can be driven.

This means that decisions can be made on forecasts which consider uncertainty in demand and uncertainty in the forecast. The forecast can then be used to drive actions downstream, such as determining optimal safety stock levels, which is an example of an action layer from forecasting.

Benefits

• Have access to more real time information from a wider variety of sources in the business, to drive actionable decisions
• Understand and use uncertainty within the forecast, which is often overlooked traditionally
• Understanding the resourcing (human) and capacity (machine) required to meet demand with supply on a frequent basis
• More accurate inventory levels which are linked to demand – ensuring there’s no over/under stocking of items
• Increased working capital as a result of optimal inventory levels and no capital tied up in stock
• Match your product ranges to meet demand at all times, ensuring you are manufacturing the goods which are needed for customers
• Increased customer service levels as a result of being able to cope with demand and ensure customers get their desired goods at minimal time frames

Pricing dynamically

Historically

Pricing often appears to customers to be fairly static in nature – either based on a list price for a product or an agreed contract price based on volume of a particular product. These are often decided at wide range intervals, either yearly or upon contract renewal, which does not necessarily consider price fluctuations during the interval.

There are also nuances where, historically, a sales representative has agreed discounted prices with customers based on past custom and the quality of their relationships. Both above examples can result in manufacturers losing out on revenue, and margin – particularly if raw material costs begin to rise, or internal production costs increase.

This is not the most optimal way to price given modern and changing customer demand, as it results in a lack of consistency in pricing which means money is often left on the table during sales and negotiations.

Using artificial intelligence to deal with disruption

Pricing should be dynamic, constantly updating to take into consideration consumer demand, market trends and internal costs. This enables manufacturers to price their products in a sweet spot based on price elasticity; a measure of how sensitive the quantity demanded is to its price.

It can also consider live price variations in raw materials to ensure that margin (if this is a key metric) can be maintained. AI is capable of driving live pricing based on multiple factors, to give recommendations in a wide range of pricing types (such as list price, live quote price, term deal pricing and  price referral automation.)

This helps to increase revenue and margin for manufacturers by ensuring products are optimally priced in line with demand, competitors and previous product prices. It takes away the need to worry about spending time manually updating prices, and ensures consistency across prices throughout their goods.

Benefits

• Increase key business metrics – maximize margin and profitable sell-through
• Simplifying a complicated manual process which typically takes considerable time to perform
• Increased consistency across pricing throughout the business as a whole
• Make intelligent pricing decisions on every SKU at more frequent intervals
• Considers internal and external factors on a regular basis to adjust pricing
• Adjust pricing for key demand changes and product level decisions

Delivering efficiently

Historically

Manufacturers are not dissimilar from any other business delivering products – their goal is to have an OTIF (On Time In Full) rate of 100%. This means that their products are delivered accurately (correct items and quantities) and on time (by the agreed date). Whilst this should be a key priority for manufacturers, it can be difficult to attain.

Products have historically been primed for delivery on a first come, first served basis based on customer order information – both from a picking and logistics perspective. Traditionally, these products are usually served out of distribution centers (either owned or third-party managed) with limited data and thought behind why products are placed in particular locations – which can have consequences for speed of delivery and customer service levels.

Both areas have not consistently and accurately used data sources (e.g. demand forecasts, production schedules, warehouse picking data, logistics guardrails) to help them deliver efficiently.

Using AI

AI can help drive supply which is matched to demand levels and production rates, to ensure customers get the right product, in the right place, at the right time. This is something which dramatically increases customer service levels, but has come to be expected in a modern world where companies such as Amazon can turn orders around in as little as 24 hours – sometimes sooner.

Examples of how AI can help with efficient delivery include…

• Sensing abnormal demand and late customer order changes (so that customers can still get their orders picked and delivered within their original timelines)
• Adjustments in manufacturing labor levels and production schedules to meet demand changes
• DRP (Distribution Requirements Planning) to ensure SKUs are sent to the correct nodes based on demand
• Wave and pick face optimizations to ensure products are picked on a timely basis to ensure delivery targets are met
• Route optimizations to allow products to be delivered via the best route for the customer and business – saving time and costs and increasing customer satisfaction. There are also considerable environmental benefits to this, which contribute to sustainable business objectives, which are being increasingly prioritized by many

Benefits

• Increased customer satisfaction levels as products are delivered accurately, at the promised date
• Increased efficiency from manufacturing lines and warehouse pickers
• Reduced manufacturing and logistics costs as products are produced and delivered accurately
• Faster delivery and turnaround times for orders compared to traditional methods

AI for manufacturing: wrapping up

The manufacturing industry has been ever evolving with the introduction of exciting new technologies. The rise of AI enables manufacturers to deliver a multitude of benefits in a dynamic world. It does not require a rip and replace introduction, and centers around driving insight from the existing data sets within the business – allowing full visibility within the supply chain with actionable decision insights to help manufacturers compete and win.

It can deliver tangible benefits in specific areas such as demand, pricing and supply, for example – but the synergies which a connected supply chain can bring to a business are huge and transformative.

Peak an AI company. We have the capability to deliver these solutions in all areas, with the unique offering that we can provide a connected intelligence layer across the full supply chain; guaranteeing that we can drive revenue, margin and service levels through winning decisions.

Drop us a line if you’d be keen to learn more.