How will AI transform the supply chain?
While machine learning software and artificial intelligence (AI) have arguably been used to reinvent marketing, a new survey of retail supply chain professionals, mostly in the grocery space, shows that only one in three have incorporated AI capabilities into their supply chain management processes. Another one in four are working toward that goal.
The survey from Symphony RetailAI, a provider of such AI-driven solutions, also found that disconnected systems, poor demand forecasting accuracy and dated technologies remain key challenges in supply chain management for high-volume retailers.
Among the findings:
- Nearly two thirds of retail supply chain professionals struggle with disconnects between systems.
- Forty-eight percent rate their forecasting technology as average to very poor.
- Forty-two percent describe less-than-optimal synchronization between their inventory and channels, with nearly as many worrying about fulfillment complexities, stocking inefficiencies and high product lead time.
- Forty-three percent said they’re challenged by lack of real-time visibility into all supply chain inventory.
AI quality and speed were seen as the two areas where the technology held its greatest potential for the supply chain. Fifty-nine percent selected either “real-time supply chain updates” or “quality control” as the supply chain processes most likely to be impacted by the integration of AI. Other predictions for AI included improved demand management (49 percent) and the ability to identify and address inefficiencies (44 percent).
AI has applications in the supply chain for everything from autonomous robots, drones, self-driving vehicles and shipment-tracking sensors to more sophisticated software that can help forecast inventory counts or transportation times.
Beyond addressing labor constraints, a data-driven supply chain supported by AI promises to support demands for mass customization and innovation, according to a new report on the topic from The Consumer Goods Forum and PA Consulting. The study states that AI and robotics automation “could become the autopilot behind a supply chain, which handles planning and fulfillment, designs products, monitors inventory levels, optimizes sourcing, synchronizes production and maintains machinery.”
- Study Shows Half of Retail Leaders Are Embracing AI to Transform the Supply Chain – Symphony RetailAI
- Strengthening the Retail Supply Chain (study) – Symphony RetailAI
- Ai And Robotics Automation In Consumer-Driven Supply Chains – Consumer Goods Forum
- Where is supply chain software headed? – Supply Chain Management Review
Discussion Questions
DISCUSSION QUESTIONS: Where do you see the greatest near-term and long-term benefits of AI in the supply chain for retailers? How do you see AI’s influence on the supply chain changing expectations about the shopping experience?
Poll
BrainTrust
Gib Bassett
Director, Solutions Marketing with Alteryx
Andrew Blatherwick
Chairman Emeritus, Relex Solutions
The article really says it all. AI can use past data to match up with current situations related to production, sales, weather related issues and more. I’ve attended IBM’s conferences on AI for the past several years and watched multiple demonstrations on the power of using AI in supply chain management. Some of the benefits include better inventory management, not running out of items, more control of just-in-time strategies at all levels (manufacturing, warehousing, stocking the retail shelves, etc.), etc. And this is just the start.
After labor costs, inventory is the most expensive part of retail. AI has tremendous potential to lower inventory costs in many ways: more accurate forecasting, customizing assortments by local market, substantially lowering out-of-stocks, and the list goes on. AI is also the real-time nerve center that makes click and collect possible, efficient and profitable. However, we must not forget what the acronym stands for – “I” is for “Intelligence.” Intelligence requires connected data, across all silos and systems. Said another way, many brick-and-mortar retailers will be challenged by the technology and costs of creating the infrastructure required to feed the AI beast with the data it requires to be “intelligent.”
I think the overall issue is potential versus reality. There is no doubt that AI is the future in many areas and the retail supply chain is just another area where it can succeed. However, there are still too many glitches that cause problems, and there is no doubt that retail supply chain professionals are skeptical because of what can go wrong. In many cases it has, and has cost significant dollars. The problem is tech companies are pushing their technology faster than its development. Tech developers are brilliant people with great skill who need time to perfect their technical advancements, but too often those driving these tech companies are selling and making promises that are not realistic. Part of the reason is that technology is moving so fast there is a great desire to be there first but delivering inferior products is not a good solution. I see the future full of new and outstanding technology with incredible conveniences, but I strongly recommend developers deliver what they know is working and be more upfront with their customers. It may not be as profitable short-term, but will definitely be long-term.
Evidence so far suggests having a corporate plan for analytics and AI first is the winning path that might land on any number of supply chain use cases. Few retailers have done this. Chasing supply chain AI use cases discretely apart from an overall plan isn’t likely to result in success. At the same time, planning and strategy do not have to be a laborious and long term activity. It does take C-level commitment and a focus on prioritizing use cases based on potential business value. I just posted about this subject.
The basic change from gut and guesswork to cloud-based AI-fueled decisions for daily store demands is a big near-term area of benefit that touches the customer, the store, and the upstream supply chain.
The benefits include more accurate forecasting models, more efficient use of store staff time with real-time access to inventory and order data, and a continuous AI learning loop using the point of sale data. This scenario makes a significant step into eliminating short shelf-life food waste, more efficient use of store labor and happy consumers finding the items they want.
AI is already transforming the supply chain for innovative retailers. Particularly in the area of demand forecasting, which affects so many other business functions, we are seeing true machine learning positively impact the accuracy of demand forecasts so inventory visibility and management are showing improved performance. It is true that only a small minority of companies are implementing these technologies today, however if more investigate the ROI once these capabilities are achieved, I believe we will see more widespread adoption.
There is massive hype about AI and machine learning at the moment. How many people really understand what it can deliver and how soon that will happen? The truth is that AI is not a product, it is already embedded in a number of modern supply chain optimization solutions giving much needed help in promotional analysis and forecasting, the impact of weather through regression analysis and many more functions within the process. Similarly, on the physical side, it will eventually bring about changes in robotics, moving them from simple machines to intelligent machines that will have a significantly bigger impact on the operations than at present.
The evolution of all software solutions has been traveling at a very fast pace for many years and AI is just the next aspect of that development. It’s a phrase coined by the consultants to add value to their work and revenue to their companies. The real issue is companies embracing the new technologies as there has been significant underinvestment in supply chain optimization solutions in recent years, with many retailers falling behind and that is affecting their ability to stay competitive.
Don’t wait for that magic day when “AI is here.” It is already a reality and solutions will continue to evolve and utilize AI to move forward enabling retailers to benefit from those developments. Sitting back waiting for the magic day will result in being out of business while you wait.
AI in the supply chain will first impact infrastructure, then customer on-demand fulfillment.
So far, supply chain has been the key place where AI fits and has shown exacting value. Companies like Blue Yonder (recently acquired by JDA) developed demand forecasting solutions for retailers with real upside and ROI. AI has only been able to fit into the supply chain so far, but at various points — from sourcing to transportation to shelves. AI impacts all predictive, data driven modeling with greater accuracy and on larger scale, can influence out-of-stocks, product availability and demand forecasting.
The real benefit of AI near term, however, will be forcing retailers to integrate disparate systems to take advantage of AI and catch up to and compete with peers or newer players who aren’t tangled with complex systems.
Long term, AI will provide retailers the ability to offer on-demand/just-in-time store availability as well as delivery across broader distribution systems including ecommerce. As AI becomes even more influential, it will expand to upstream and downstream relationships between manufacturers, suppliers, and retailers, where it will be able to adjust a manufacturer’s output or raw material sources to provide accurate fulfillment of goods at the store three months out. The shopping experience will change with it, both improving and spurring on further customer on-demand expectations.
As the speed and complexity of supply chain requirements increase due to faster speed and flexibility demanded by the consumer, AI will be a tool that helps reduce inefficiencies and reduce costs. However, the biggest challenge for AI is having accurate and sufficient real-time data for AI to make good recommendations. Before embarking on AI solutions, the user of a retail supply chain needs to understand the AI objectives, processes being measured and he the required data is to be collected. AI is not magic, and these factors must all be considered.
As AI provides for increased efficiencies and lower cost, consumers expectations for speedy and accurate delivery will go up. This will put increased demands on the retail supply chain.