Traditional automation technologies alone are not sufficient to enable the fully automated operation of trains. However, Artificial Intelligence (AI) and Machine Learning (ML) offers great potential to realize the mandatory novel functions to replace the tasks of a human train driver, such as obstacle detection on the tracks. The problem, which still remains unresolved, is to find a practical way to link AI/ML techniques with the requirements and approval processes that are applied in the railway domain. The safe.trAIn project aims to lay the foundation for the safe use of AI/ML to achieve the driverless operation of rail vehicles and thus addresses this key technological challenge hindering the adoption of unmanned rail transport. The project goals are to develop guidelines and methods for the reliable engineering and safety assurance of ML in the railway domain. Therefore, the project investigates methods to reliable design ML models and to prove the trustworthiness of AI-based functions taking robustness, uncertainty, and transparency aspects of the AI/ML model into account.

A characteristic of reinforcement learning is the ability to develop unforeseen strategies when solving problems. While such strategies sometimes yield superior performance, they may also result in undesired or even dangerous behavior. In industrial scenarios, a system's behavior also needs to be predictable and lie within defined ranges. To enable the agents to learn (how) to align with a given specification, this paper proposes to explicitly transfer functional and non-functional requirements into shaped rewards. Experiments are carried out on the smart factory, a multi-agent environment modeling an industrial lot-size-one production facility, with up to eight agents and different multi-agent reinforcement learning algorithms. Results indicate that compliance with functional and non-functional constraints can be achieved by the proposed approach.

Abstract Systems are becoming increasingly more adaptive, using techniques like machine learning to enhance their behavior on their own rather than only through human developers programming them. We analyze the impact the advent of these new techniques has on the discipline of rigorous software engineering, especially on the issue of quality assurance. To this end, we provide a general description of the processes related to machine learning and embed them into a formal framework for the analysis of adaptivity, recognizing that to test an adaptive system a new approach to adaptive testing is necessary. We introduce scenario coevolution as a design pattern describing how system and test can work as antagonists in the process of software evolution. While the general pattern applies to large-scale processes (including human developers further augmenting the system), we show all techniques on a smaller-scale example of an agent navigating a simple smart factory. We point out new aspects in software engineering for adaptive systems that may be tackled naturally using scenario coevolution. This work is a substantially extended take on Gabor et al. (International symposium on leveraging applications of formal methods, Springer, pp 137–154, 2018).

Short product lifecycles and a high variety of products force industrial manufacturing processes to change frequently. Due to the manual approach of many quality analysis techniques, they can significantly slow down adaption processes of production systems or make production unprofitable. Therefore, automating them can be a key technology for keeping pace with market demand of the future. The methodology presented here aims at a meta-model supporting automation for PFMEA. The method differentiates product requirements, production steps and quality measures in such a way, that complex quality requirements can be addressed in any instance of a factory using a common meta-modeling language.

Adversarial learning has been established as a successful paradigm in reinforcement learning. We propose a hybrid adversarial learner where a reinforcement learning agent tries to solve a problem while an evolutionary algorithm tries to find problem instances that are hard to solve for the current expertise of the agent, causing the intelligent agent to co-evolve with a set of test instances or scenarios. We apply this setup, called scenario co-evolution, to a simulated smart factory problem that combines task scheduling with navigation of a grid world. We show that the so trained agent outperforms conventional reinforcement learning. We also show that the scenarios evolved this way can provide useful test cases for the evaluation of any (however trained) agent.

According to statistics and future prospects in the next few years world-wide energy consumption will increase significantly. Therefore not only more energy efficient technologies but also more extensive energy saving concepts have to be realized. We have developed an rdquoimplicit interactionrdquo based power saving concept, which automatically schedules and controls energy consumers depending on the recognized activities of users. Moreover, dynamically schedulable power consumption loads are shifted in time, so as to make use of the cheapest energy prices without compromising user comfort. In order to calculate the potential savings of this concept we have implemented a simulation framework executing energy consumption models on a city scale, which allows for more complex scenarios than being restricted to some devices or buildings only. Both the framework architecture as well as large scale energy consumption simulations are presented. Simulation experiments give strong evidence for our implicit energy management concept as a promising source of energy savings.