Accessibility Links

Supply Chain Management in 2019: The Impact of AI

Date: 11 February 2019


Artificial Intelligence is becoming more commonplace in people’s daily lives than ever before. From ‘smart replies’ in email at the click of a button (‘Thanks!’, ‘Let’s do it!’) in response to your email content, chatbots so accurate you think you’re speaking with a real person, to Google’s algorithm of course, using machine learning from previous searches to show you the ‘best answer’ to your question - AI is everywhere.

Many companies are starting to apply artificial intelligence (AI) across global supply chain management to improve efficiency as well as speed up or even automate decision making.

The opportunities AI brings, such as reducing the cost of human labour, makes it highly appealing to many industries globally. But is it the right thing to do?

In this article we look at problems of the supply chain, possible AI solutions, and how desirable it is to implement them. 

Supply Chain Challenges

Since the beginning of commerce and manufacturing, all supply chains have a similar goal: to move goods or services efficiently through the system without bottlenecks, overstock, or undersupply. One of the most persistent issues in achieving this is rising costs. 

“Appraisal costs, energy, operating fuel, transportation – essentially anything to do with moving goods through the chain  – presents an increasing financial challenge. Growth in the number of global customers, along with the rising costs of labor and commodity, all put a strain on a company’s bottom line,” notes Nick Aoun, Head of Services, Principal Product and Supply Chain Consultant at DSJ Global. 

Another major problem is customer service: the right product being delivered in the right quantity at the right time. “This can get complicated quickly when you have different stakeholders involved. Human error is almost inevitable”. 

AI Solutions for Supply Chain Problems

Advances in communication and technology have always been a huge catalyst for change. As we move towards artificial intelligence, machine learning, ‘Internet 4.0’ -  essentially where machines communicate with each other and make decisions a human would usually make, there is are opportunities for major innovation in the supply chain. 

Three main areas where AI may most effectively be implemented are: 
Predictive analytics
Warehouse management 
Procurement via chatbots 

Predictive Analytics

Predictive analytics is a type of machine learning/artificial intelligence where large data sets are mined for information that can be used strategically. Predictive analytics could help manufacturers set prices, accurately forecast demand, and even prevent disruptions in the supply chain by scheduling repairs in anticipation of a breakdown.

One of the most advanced examples of AI usage in predictive analytics is IBM’s Watson Supply Chain, which gathers and interprets internal data in addition to data from weather, news, and social media sources. Supply chain technology provider Lenovo has partnered with Watson and claims to have helped them find problems up to 95% faster, as well as identify opportunities to lower costs and increase revenue.

Predictive analytics is therefore already a reality and may soon be suitable for widespread implementation. 

Warehouse Management

Artificial intelligence appears to be a good fit for warehouse management systems (WMS) which are responsible for smooth operations at company facilities. Swisslog, a logistics company based in Switzerland, is touting the development of a “learning warehouse,” which will be integrated into its buildings. It uses AI to automate the entire workflow and human-machine interaction.

However, this technology seems a bit farther out; AI research company Emerj found no evidence of Swisslog’s learning warehouse in action.

“Predictive analytics and WMS is still a very new topic. It’s not something that’s been perfected in any way, shape or form,” Aoun commented. “A McKinsey report said Google and Baidu spend $20 - $30 billion on AI last year, with 90% on R&D and the rest on acquisition.”

Chatbots

Chatbots can be installed quickly and quickly add transparency and efficiency to the procurement process.
One of the areas that appears the most ready for implementation is the chatbot, which can interact with all players in the supply chain, providing real-time updates, answering frequently asked questions, sending compliance or government notifications, and more. Eventually, these chatbots may be able to negotiate prices or handle more complex matters.

How feasible are AI implementations for the supply chain? 

For supply chain management, the reality is AI is still in its earliest stages. According to research by HBR, AI SCM is only being used by 7% of manufacturing and supply companies to automate production - but it looks set to increase. 

Something which may be holding back widespread adoption of AI is the lack of tech talent to make these things happen. The big players (Google, Baidu) have cornered the market on the technology and pay top dollar to their highly skilled development teams. 

“AI specialists are so few and far between. To find someone who will do NLP (natural language processing) even for a tier-one automaker in Germany or France, for example, is very difficult,” Aoun says. 

The Future 

Further down the road, we may see widespread use of AI, with software and hardware eventually replacing all human roles along the supply chain. A 2013 Oxford University study predicted 47% of jobs could be automated by 2033.
 
But is this the right thing - both ethically and economically for society? As the statistics show, unemployment creates a huge range of problems. Homelessness, civil unrest, crime, reliance on state benefits are all serious concerns when unemployment levels rise. 

Perhaps caution is needed when it comes to AI implementation. 
“What companies should do is move slowly – have humans lead the robots, not the other way around.’ suggests Aoun. 

Are you looking for supply chain talent to help grow your business?