Admittedly, Artificial Intelligence (AI) is a nebulous term – there is nothing out there that is truly AI at this point. When most people talk about AI they are talking about Machine Learning (ML). Machine Learning is much more straightforward in that it involves training a machine by giving it samples of labeled real-world data, including text and visual inputs. By feeding all this into the machine you can train it on what things are. For example, after training a machine with hundreds of images of cars, it will be able to classify a new image of a car as a car.
There are 4 types of machine learning: Supervised learning, Unsupervised learning, Semi-Supervised learning, and Reinforced learning. In supervised learning, the machines are trained using labeled datasets. Machines use this information to predict output in the future. In Unsupervised learning, unlabeled and unclassified datasets are used to train the machines. They then predict the output without supervision or human intervention. This method is commonly used to categorize unsorted data based on their features, similarities, and differences. Machines are also capable of finding hidden patterns and trends from the input. Semi-supervised training is a combination of labeled and unlabeled datasets being used to prepare the machines. And lastly, in reinforcement learning, there is no concept of labeled data – machines learn only from experiences. Using a trial and error method, learning works on a feedback-based process.
ML requires a lot of computing, storage, and data. With the availability of cloud computing and Graphical Processing Units (GPU) this became much more feasible. Deep Learning, which became very popular starting in 2012 revolutionized AI and is based on CNN (convolution neural nets) The rate of change or advancement in neural nets since then has been massive. With these developments, AI/ML has been democratized and more widely available to everyday people.
Ultimately everyone uses AI/ML whether they realize it or not, whether it’s in their car, with their Alexa, or on their phone. And of course with the publicity ChatGPT has received lately, a number of people may be trying the latest form of AI that is based on Large Language Models (LLMs). They are typically trained on large text datasets and can then be used to generate, summarize, translate or perform other language-related tasks. As a fun aside I asked ChatGPT to write me a blog on AI & blockchain in supply chain management – I’m including the result at the end of this blog. It didn’t do a bad job!
Still, I’m not ready to give the reins completely over to a machine just yet, so here’s my human brain’s take on the value of AI and blockchain as it relates to supply chains:
The integration of blockchain and AI can revolutionize supply chain and logistics operations by improving efficiency, transparency, and decision-making. AI can provide intelligent insights and data-driven analysis, while blockchain secures and tracks the data, ensuring its accuracy and integrity. Together, these technologies can enhance inventory management, route optimization, provenance tracking, and end-to-end visibility across the supply chain. By leveraging AI and blockchain, organizations can improve decision-making and optimize supply chain operations, leading to increased customer satisfaction and business success.
The Maverix AI/ML predictive analytical module is being developed to analyze vast amounts of supply chain data. Utilizing various machine learning algorithms, it will analyze historical data to identify trends and patterns, make predictions on future supply chain performance, and detect areas for improvement in the supply chain. Further, by monitoring real-time supply chain data from IoT devices (as we are doing), the module can predict potential disruptions and provide valuable insights to optimize supply chain efficiency and drive business growth.
To enhance the robustness of the AI/ML module, WaveDancer will be augmenting our customer’s supply chain data with additional open-source data such as weather and air pressure data from specific geographic regions.
Initially, the Maverix AI/ML module will focus on two main use cases:
Inventory Management: This use case will use data analysis to predict future demand and help organizations determine the optimal inventory levels, avoiding overstocking or stock shortages. The module’s real-time monitoring capabilities will allow for quick responses to changes in demand. It will use machine learning methods such as decision trees and deep learning algorithms.
Predictive Maintenance: This use case will use various algorithms such as few-shot learning algorithms, clustering algorithms, and deep learning algorithms to predict the probability of failure of different parts. The few-shot learning algorithms are designed to learn and make predictions based on limited data and are beneficial for quickly adapting to new tasks and classes.
The Maverix AI/ML predictive analytical module holds great promise for significantly improving supply chain efficiency and performance, providing valuable insights, and driving business growth.
Maverix is built on top of a private permissioned blockchain to support complex supply chains. It has been thoughtfully designed with security as its foundation, ensuring a secure experience for users.
Maverix utilizes multilayer encryption, federation of data, identity and access management, role-based access controls and rigorous testing as the standard in protection throughout the platform. This begins with the private, permissioned blockchain which supports multiple users and nodes, but allows a high degree of access control to the organization. To create an immutable data ecosystem all events will be stored on the blockchain and will be used for credentialling as well as auditing to create a multilayered approach to security. All data written to chain is obfuscated and encrypted. All personally identifiable information is stored off chain to allow conformity to global privacy standards.
Maverix’s infrastructure is such that it is scalable both horizontally and vertically to support hundreds of concurrent users. The ability to scale rapidly improves the overall user experience through speed of access to data and overall performance of the system.
Role Based Access Controls are implemented at every layer of the application through encrypted access tokens. This obfuscation of user information protects user credentialing and identity while roles and permissions limit what users have access to.
Maverix uses Smart Contracts for autonomous administrative functions, reducing cost and potential human error while increasing data integrity. This includes but is not limited to thresholds for acceptability as it relates to the timeliness of assets, invoice reconciliation, and governance.
ChatGPT’s Output for “The value of AI & blockchain in SCM”:
Artificial intelligence (AI) and blockchain are two of the most disruptive technologies of the 21st century, and their combination has the potential to transform supply chain management in ways that were previously thought to be impossible.
The supply chain is a complex system that involves the movement of goods and services from the raw materials stage to the end consumer. It is a critical component of the global economy, but it is also plagued by inefficiencies and challenges, such as lack of transparency, fraud, and inefficiencies.
AI and blockchain offer new and innovative solutions to these problems. AI can be used to analyze large amounts of data, identify trends and patterns, and make predictions about future events. This allows companies to make more informed decisions, improve operational efficiency, and reduce waste. For example, AI can be used to optimize the routing and scheduling of trucks, reducing the time and cost associated with transportation.
Blockchain, on the other hand, offers a new level of transparency and security to supply chain management. It is a decentralized, tamper-proof ledger that records every transaction in a secure and transparent manner. This allows companies to track the movement of goods and services in real-time, reducing the risk of fraud and improving the overall efficiency of the supply chain.
When combined, AI and blockchain have the potential to revolutionize supply chain management. For example, blockchain can be used to create a tamper-proof, decentralized database of all supply chain transactions, and AI can be used to analyze the data in real-time, providing insights into the supply chain that were previously impossible to obtain.
In conclusion, AI and blockchain are powerful technologies that are already transforming supply chain management. By improving transparency, security, and efficiency, these technologies have the potential to create a new, more efficient and more secure supply chain that benefits everyone involved. As these technologies continue to evolve, it is likely that we will see even more innovative solutions emerge, and it will be exciting to see how they will continue to shape the future of supply chain management.