Artificial intelligence in supply chain management : Top 7 trends in 2020

-

The marriage between AI and supply chain management

In this article I attempt to provide a brief snapshot of  how the AI-powered systems are driving the supply chain management field in 2020. These may include trends that are impactful breakthrough but still maturing as a sustainable solution  as well as those in the post plateau and has matured into growth phase. 

The area of dynamic and intelligent approach to supply chain management has drawn immense attention in later half of last decade.The explosive boom in the AI sector has undeniably cast a spell on the way supply chain and logistics industry has been conducting its processes.

In the following list, along with the trend, I also mention a selecta of players in the particular technology/trend, and the way they are making an impact.

                                                 So here we go!

1. Process automation  
Robotic process automation (RPA) enables automation of manual or human involved business process, by means of intelligent and autonomous software robots. Uipath a major player in this area, provide a comprehensive automation suite for end-to-end supply chain process optimization. In the physical automation arena, autonomous vehicles in warehouse automation and drones in same day delivery are ruling the roost. Amazon has already revealed their first version in last mile delivery innovation. 

2. Multi-agent based optimization 
 Intelligent agents have adaptive learning capacity which could be leveraged to make decisions in a collaborative environment like resource allocation,distribution planning and scheduling or a competitive environment like bidding e-auctioning This is because they could be trained to have both self and collective awareness for decision making.The difference here from RPA is principally rule based routine logic enabled automation whereas multi-agents can learn and improvise.

3. Meta learning 
 Meta learning is essentially "learning how to learn" .In particular reinforcement learning is earning attention in learning about the environment dynamically and continuously improving on solutions based on stochastic semi-markov or markov decision process modelling.Meta reinforcement learning is already a trend in Google's AI platform like Deepmind .

4. Blockchains 
 Blockchains offer provenance enabling digital peer-peer traceability ecosystem.In food and agribusiness supply chains blockchains are already making an impact.But other business environments also invite further expansive empirical testing of this technology. According to Consensys (an ethereum based developer platform) blockchain facilitates disintermediation of the supply chain network.On the other hand, Origintrail offers decentralized data network that intermediates data exchange between blockchain and web based user applications providing knowledge-graph based interoperability.

5. IoT cloud based infrastructure 
 IoT based PaaS and SaaS are revolutionizing the speed and efficacy of supply chain management through enabling digitized ecosystem.One huge spin-off from this venture is improved end to end visibility solutions. IBM and SAP leads the frontier in PaaS or SaaS solutions for enterprise level management of end-to-end supply chain process visibility and optimization.Both offer innovative customizable solutions for a digital supply chain.RFID enabled data logs and event monitoring is at the core of the IoT ecosystem.

6. Digital Twins  
According to Glaessegen and Stargel (2012) ‘digital-twin are integrated multiphysics, multiscale, probabilistic simulation of a complex product, which functions to mirror the life of its corresponding twin’.
Siemens and GE have already made pioneering advances in digital-twin enabled Product lifecycle management (PLM).Supply chain data can be tracked and transformed into value-added assets, like exemplified by these players and many others. Digital healthcare supply chain can also hugely benefited with designing patient-centric analytics based on digital twins.It can also enable an immersive customer experience by managing the product lifecycle and providing aftermarket support.

7. Predictive and Prescriptive Supply chain Analytics
According to INFORMS analytics book of knowledge, 
A predictive analytics model " seek to forecast the likely future state of the world through a deeper understanding of the relationships among data inputs and outcomes."
Whereas a prescriptive analytics solution "seek to go further than forecasting a future state, to make actionable recommendations about what the decision-maker should do to achieve a particular objective, such as maximize profit. Prescriptive analytics typically require a combination of simulation and optimization". 
While developing a predictive model, understanding the context ,the actors and business requirement of end user of the visualization is very crucial. Agent based simulation is a very trendy approach in this area but it requires more empirical and case based validations as compared to others like discrete-event and system dynamics. Prescriptive analytics caters to guide the decision maker through strategic and tactical supply chain decisions like where to locate the DC or type of product portfolio that's unique to customer. This article offers a good overview of how the two approaches can be managed to create an integrated and intelligent view of supply chain processes.

In my article the intelligent supply chain, I go further into details of what can constitute a digital, resilient and agile supply chain.
In upcoming post, I shall dive deeper into the area of automation in supply chains and critique the challenges and limitations.



Comments

Post a Comment

Popular posts from this blog

Becoming a Leader : A growth based maturity model

-
Image
The matter of leadership has enchanted possibly every individual, in his/her trail for excellence.Most of time, its seen as an over-night event of transformation that magically results from a spontaneous recipe of a few popular ingredients like courage, confidence and favourable externalities.People need to realize that sustainable and emulatory leadership goes much deeper than that level.I strongly feel there is a need for life-time value based maturity model of leadership, as a pathway which could be structured from very basic steps,so that it is accessible to a much wider audience.In the model that follows, I propose six phases of growth of individual leadership as a maturity mode l. These phases are as follows : Phase I :Instinctual bootstrapper Trust your gut-instinct and take the first step. The first steps land the first acquaintances Acquaintances turn into relationships through active community-level interactions Communication is key driver post-entry Phase II: Directed enthus
Read more

Value-added data-asset lifecycle management (Part II) : Enterprise capability development

-
                              Welcome back to the expository series on data assets and data governance! In my previous article, we learned about v alue-added data assets   and how and why you have to manage them strategically.In this episode, I discuss how these assets can be transformed for competitive enterprise capability development. Before delving into capability development from data assets   management, an important distinction claim attention here and needs to be delineated. That is between info quality and data quality , even though the two are used interchangeably , the former has user quality (known as information consumer) as its focus .The first major strategy shift required here is viewing information as a product to be managed as against limiting to data quality in databases and data-warehouses. Two models are considered industry standard for structuring the quality assurance which is the forerunner to information product management.They aid to assess info quality base
Read more

Blockchain, multi-agents and dynamic learning in supply chains

-
Image
Most of us in the supply chain professional community would not probably escape a day without noticing a story of how the blockchains are the talk of the 'supply chain town', with their ambitious claim to become the most promising step towards attaining the unvanquishable double T's of an agile supply chain network vis a vis traceability and transparency.There are two other wizkids of AI that promise to craft a more dynamic element to the plotline-:  intelligent multi-agents and dynamic learning . Here, I try to sketch a thread on possible interlude among these 3 game-changers in a relatively less explored context. Recently I was going through a research paper which was addressing the problem of supply chain disruption management with multi-agent based framework.Given that the treatment of problem was indeed , I found that the approach was silo-type with agents coordinating across enterprise legacy systems .The entire risk identification approach was centered on KPI benchma
Read more