Manufacturing, and the supply chains that support it, are the backbone of the global economy. They are made of countless moving parts, all of which must be orchestrated with extraordinary precision in order to ensure that everything works efficiently regardless of a changing world. As 2020 has made abundantly clear, all that we rely on can change in the blink of an eye, and with it the tried and tested ways of managing this complexity. As the 21st century progresses, the ability to adapt safely without the inherent brittleness of changing a large, complex organisation, will be increasingly the difference between operating well or not.
Autonomous Economic Agents provide autonomy to all of the individual moving parts across manufacturing and supply chains. They are able to work amongst themselves to solve problems, and do so in a hyper-local environment. Free of the complexity of a centralised single entity, these decentralised solutions are scalable, secure, non-brittle and can adapt in real-time to global and local events.
We are completely dependent on these networks, and top-down structures are strained by having to understand the entire scope of what they are and are not doing. This new approach provides a solution that can be introduced safely in parallel to existing systems. In this article, we discuss these benefits and how Autonomous Economic Agents will be a key part in delivering this vision.
Problems associated with centralized manufacturing
In particular, operation and supply chain experts contend that:
- Centralized manufacturing control systems are insufficiently flexible to respond to changing production styles and highly dynamic variations in product requirements.
- Lack of scalability, agility, and reconfiguration capabilities of manufacturing systems.
- A silo of information that leads to operational inefficiencies, as information isn’t distributed or isn’t distributed on time.
- Exposure of the system to a single point of failure, increased response overheads and handling of confidential information.
Autonomous economic agent technology provides a natural way to overcome such problems, by providing digital agency to the individual components so that they can deal with events without requiring many levels of management to make what are, in many cases, trivial local events. This can be thought of as an intelligent manufacturing environment: there is no loss of control or instrumentation, but the day-to-day minutiae of getting things done and reacting to changes can be done digitally, and free of human intervention.
Agent-Based approaches for intelligent manufacturing
Festo has been in the manufacturing sector for nearly a century. It is working with Fetch.ai to develop a decentralized manufacturing marketplace. This takes a multi-agent approach of creating digital software agents to represent all the component parts and allowing them to act on their behalf. This utilises Fetch.ai’s Autonomous Economic Agents to safely transform existing manufacturing control systems at Festo with a goal to increase efficiency, run-time discovery of new optimisations, and delivering a robust, scalable solution for the 21st century and beyond.
There are advantages of this approach, which include:
- An agent based manufacturing control system could instantaneously recognize demand fluctuations, uneven utilization of manufacturing resources, and detect inefficiencies in manufacturing capacities.
- An information network that democratizes access to data. As data becomes available from both within and outside the organization, it allows for A.I. to access a larger data set, leading to faster, more accurate, and better decision making.
- Negotiation between agents enables direct communication and interaction with buyers, products, machinery, allows for seamless automation and increases the responsiveness of each party in the supply chain. This also means that buyers can plan and predict their needs and automatically make purchases without delay. Simultaneously, manufacturers become aware of demand and capacity and can make immediate shifts in production.
Our vision for a Decentralized Marketplace for Manufacturers
In addition to leveraging Fetch.ai’s autonomous economic agents within the system, Festo also aims to create a decentralized marketplace for manufacturers where information can be bought and sold from one organization to another.
Within a centralized marketplace, platform providers can monopolize their own data and this leads to unfair pricing practices whereas a decentralized marketplace creates a highway of information transfer that any manufacturer can take advantage of and interact directly with each other thereby automating and optimizing key manufacturing processes such as planning and purchasing from external manufacturers and suppliers.
Beyond making everything “smarter,” Fetch.ai and Festo believe agent based solutions enabling the creation of autonomous, decentralized marketplaces have significant tangible business benefits. A decentralized marketplace would not only enable further optimization of existing supply chain protocols but also increase exponential cost savings associated with it.
We live in an increasingly complex digital world. The decentralised approach provides a new set of foundations as this complexity increases. It enables better insights, it couples tightly with modern machine learning and AI, and it delivers the ability to safely make changes without fear of system failure: it’s a large collection of simple things, with the autonomy to make decisions, rather than fewer complex things.
You wouldn’t take a bus to a go-karting track, as it isn’t agile enough to deal with its environment. This new approach transforms our ability to remain flexible in a world that can, and has, changed entirely overnight. Fetch.ai is proud to be working with Festo to deliver this to manufacturing and the supply chains that drive it.