Artificial Intelligence and Future Supply Chains
“Every aspect of learning or any other feature of intelligence can in principle be so...
By Pierfrancesco Manenti, vice president research, SCM World
“Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it.”
These were the words used in 1955 to launch the very first research project that coined the term ‘artificial intelligence’.
Fast forward 60 years and artificial intelligence – or machine learning as many call it – is emerging as the next big technology. 2016 has seen a race for artificial intelligence, with a number of acquisitions and large technology vendors – of the likes of IBM, Google and Amazon – launching new artificial intelligence-enabled products.
In SCM World’s 2016 Future of Supply Chain Survey, we found big jumps in importance for a series of disruptive technologies with respect to supply chain strategies, some of which were considered largely irrelevant just a couple of years ago.
One of these is machine learning, which in 2016 cemented its place in the technology mainstream. 47% of supply chain leaders from our larger community believe that artificial intelligence is disruptive and important with respect to supply chain strategies. The technology grew so rapidly in importance over the last couple of years that in 2014 it wasn’t even included in the research!
What’s Artificial Intelligence?
Artificial intelligence can be defined as the use of computers to simulate human intelligence, specifically including learning – the acquisition and classification of information, and reasoning – finding insights into the data. At the core of artificial intelligence is the ability to recognize patterns across the 3Vs of big data (volume, velocity and variety) and find correlations among diverse data.
Today, the term artificial intelligence encompasses everything from speech recognition to machine vision and from chatbots to collaborative robotics. The benefits of this technology lie in speed and accuracy beyond the reach of human capabilities, which is also feeding a debate about its implications in the future of work.
Business activities that require to collect and analyze lots of structured and unstructured data can benefit from artificial intelligence and its ability to support faster and smarter decision making. Supply chain is therefore a natural fit for artificial intelligence.
Artificial Intelligence and Supply Chain
An interesting 2010 research paper from Dr. Hokey Min from the College of Business at Bowling Green State University, predicted a number of applications of artificial intelligence in supply chain management. These include setting inventory safety levels, transportation network design, purchasing and supply management, and demand planning and forecasting.
Today’s artificial intelligence is mature enough to make some of those applications possible:
- Capitalizing on the machine-learning capabilities of IBM’s Watson, IBM has recently launched Watson Supply Chain aimed at creating supply chain visibility and gaining supply risk insights. The system uses cognitive technology to track and predict supply chain disruptions based on gathering and correlating external data from disparate sources such as social media, newsfeeds, weather forecasts and historical data.
- ToolsGroup’s supply chain optimization software is rooted in machine-learning technology. One area of application is new product introduction. The software begins with creating a baseline forecast for the new product. As the algorithm learns from early sell-in and sell-out demand signals, it layers this output to determine more accurate demand behavior, which feeds through to optimized inventory levels and replenishment plans.
- The machine-learning technology of TransVoyant is able to collect and analyze one trillion events each day from sensors, satellites, radar, video cameras and smartphones. In logistics applications, its algorithm tracks the real-time movement of shipments and calculates their estimated time of arrival, factoring the impact of weather conditions, port congestion and natural disasters.
- The technology firm Sentient uses machine learning to deliver purchasing recommendations to e-commerce shoppers based on image recognition. Rather than only using text searches and attributes like color or brand, the software find visual correlations with the items that the shopper is currently browsing through visual pattern matching.
- At the core of Rethink Robotics’ collaborative robots is an artificial intelligent software that allows the robot to perceive the environment around it and behave in a way that’s safe, smart and collaborative for humans working alongside production lines.
The awareness and ability to make fact-based decisions that artificial intelligence makes possible is completely new to supply chain management. This technology is expected to create the sentient supply chain of the future – able to feel, perceive and react to situations at an extraordinarily granular level.
Welcome to the future!
The January issue of Supply Chain Digital is live!
Follow @SupplyChainD on Twitter.
5 Minutes With: Jim Bureau, CEO Jaggaer
What is data analytics, and why is it important for organisations to utilise?
Data analytics is the process of collecting, cleansing, transforming and analysing an organisation’s information to identify trends and extract meaningful insights to solve problems.
The main benefit for procurement teams that adopt analytics is that they’re equipped to make faster, more proactive and effective decisions. Spend analysis and other advanced statistical analyses eliminate the guesswork and reactivity common with spreadsheets and other manual approaches and drive greater efficiency and value.
As procurement continues to play a central role in organisational success, adopting analytics is critical for improving operations, meeting and achieving key performance indicators, reducing staff burnout, gaining valuable market intelligence and protecting the bottom line.
How can organisations use procurement analytics to benefit their operations?
Teams can leverage data analytics to tangibly improve performance across all procurement activities - identifying new savings opportunities, getting a consolidated view of spend, understanding the right time for contract re-negotiations, and which suppliers to tap when prioritising and segmenting suppliers, assessing and addressing supply chain risk and more.
Procurement can ultimately create a more comprehensive sourcing process that invites more suppliers to the table and gets even more granular about cost drivers and other criteria.
"The main benefit for procurement teams that adopt analytics is that they’re equipped to make faster, more proactive and effective decisions"
Procurement analytics can provide critical insight for spend management, category management, supplier contracts and negotiations, strategic sourcing, spend forecasting and more. Unilever, for example, used actionable insight from spend analysis to optimise spending, sourcing, and contract negotiations for an especially unpredictable industry such as transport and logistics.
Whether a team needs to figure out ways to retain cash, further diversify its supply base, or deliver value on sustainability, innovation or diversity initiatives, analytics can help procurement deliver on organisational needs.
How is data analytics used in supply chain and procurement?
Data analytics encompasses descriptive, diagnostic, predictive and prescriptive data.
Descriptive shows what’s happened in the past, while diagnostic analytics surface answers to ‘why’ those previous events happened.
This clear view into procurement operations and trends lays the groundwork for predictive analytics, which forecasts future events, and prescriptive analytics, which recommends the best actions for teams to take based on those predictions.
Teams can leverage all four types of analytics to gain visibility across the supply chain and identify optimisation and value generating opportunities.
Take on-time delivery (OTD) as an example. Predictive analytics are identifying the probability of whether an order will be delivered on time even before its placed, based on previous events. Combined with recommendation engines that suggest improvement actions, the analytics enable teams to proactively mitigate risk of late deliveries, such as through spreading an order over a second or third source of supply.
Advanced analytics is a research and development focus for JAGGAER, and we expect procurement’s ability to leverage AI to become even stronger and more impactful.