The mission of the Local4Global project is to develop and extensively test and evaluate in real-life TSoS, a generic, integrated and fully-functional methodology/system for TSoS, with the following atributes:
- The TSoS constituent systems are operating as fully autonomous units that react and interact depending only on their local environment in order to optimize the TSoS emerging performance at the global level
- There will be no need for an elaborate and tedious effort to deploy the Local4Global system or to re-design/re-configure it in cases of changes in the topology, environment or hierarchy of the TSoS. In essence, the Local4Global methodology will provide a “plug-and-play control mechanism” for the constituent systems with the ability to fully exploit each constituent system’s abilities by embedding within it learning, evolving and self-organizing capabilities
- Moreover, there will be no need for an elaborate, “expensive” infrastructure that provides each and every constituent system with information coming from all over the TSoS.
The Local4Global methodology/system will be applicable to generic TSoS that comprise highly heterogeneous TSoS. Moreover, it will, by its very nature, be totally scalable and computationally efficient.
Local4Global Key Approach
The key idea for meeting the ambitious objectives of Local4Global is by suitably generalizing the well-known in control theory Certainty Equivalence Principle to the case of TSoS (see Figure 1 for a schematic diagram; see also section 1.2.1 for a more detailed description):
- Within each of the constituent systems, a self-learning mechanism is embedded which provides a “just enough” estimate of the TSoS dynamics. Moreover, a situation awareness mechanism is responsible for extracting – from local measurements – the absolutely necessary global TSoS information needed in order for the constituent system to calculate optimal actions.
- The information provided by the self-learning and the situation awareness mechanisms are used by a distributed optimizer in order to calculate the constituent system optimal actions.
These actions are appropriately modified by the so-called Control for Learning and Learning to Control (C4L/L2C) mechanism, so as the constituent systems’ actions concurrently attempt to optimize the TSoS performance at the global level and maximize the learning capabilities and situation awareness of the constituent systems. In essence, the C4L/L2C mechanism compensates for “mistakes” made by the self-learning and situation awareness mechanisms and guarantees the efficient performance of the overall Local4Global system
Progress beyond the state-of-the-art required
- Scalable and efficient Distributed Optimization tools are required for highly heterogeneous TSoS with complex and, sometimes, frequently changing topology, hierarchy and number of constituent systems, as well as constantly changing access to available information.
- Development of self-learning and situation awareness mechanisms capable of providing with “just enough” learning and information for complex and constantly changing TSoS dynamics.
- Appropriately extend Local4Global ingredients for the self-learning and self-tuning of complex control systems towards the development of the Control for Learning and Learning to Control (C4L/L2C) mechanism.
Embed Local4Global with the abilities to identify and predict the emerging/evolutionary characteristics at the macro-level and control the overall TSoS emergent/evolutionary behaviour in the “right direction”.
Integration, Use Cases and Final Product
All the above advances will lead to a fully-functional and ready-to-use system (Local4Global final product) – delivered in the form of an embedded, web-based, “plug-and-play” software system for generic TSoS, mountable locally to each constituent system. This system will be deployed and extensively tested and evaluated in 2 real-life TSoS Use Cases, a Traffic TSoS Use Case and an Efficient Building TSoS Use Case.