Search

The design of real-time systems can be accomplished using various modelling techniques in which the timing specification has to be included. At such level of abstraction, the functionality might be described using real-time procedures. These procedures specify the actions to be accomplished without any knowledge about the system architecture, in terms of nodes and network. When a procedure maps to various tasks, the determination of the tasks’ and messages’ parameters is not trivial. The technique presented in this paper builds upon previous work by the authors and helps to reduce the time from design to implementation in real-time systems.

Autonomous Mobile Robots (AMRs) are becoming pervasive and, in consequence, the optimization of their electronics is also becoming a current research topic. In particular, the use of distributed architectures is an issue. In the presence of distribution, a network must be used to interconnect the several modules of an AMR. This network must be able to support the traffic requirements to carry on the data flows produced by the nodes. In this paper a characterization of the subsystems required in AMRs is made, analyzing typical AMRs. Then, the data flows produced by them are quantified and the adequacy of popular fieldbuses, in particular CAN - Controller Area Network, to accommodate the required traffic in two robots is analyzed.

In this paper a simulator to preview the timeliness of the transmission of messages and of the execution of tasks in a distributed system is presented. The simulator, called SIMHOL, builds on previous work by the authors in which a simple mechanism to dispatch tasks and messages was proposed for CAN-based distributed systems. The inputs to the simulator are the so-called data streams, which include the producer tasks, the correspondent messages and the tasks that use the transmitted data. Using the worst-case execution time and transmission time, the simulator is able to verify if deadlines are fulfilled in every node of the system and in the network. Besides discussing the simulator construction and operation, the paper presents some examples of distributed systems requirements and the results obtained by using the tool to analyze the respective timeliness. This is also used to illustrate the outputs of the simulator. One important issue also discussed is the easy way, due to the object oriented approach chosen, to extend the simulation to cover different networks and to use different scheduling techniques either at the task level or at the message level.

The amount of information that must be exchanged in modern Digital Computer. Control Systems (DCCS) and real-time industrial automation is now reaching the limits that are achievable with traditional fieldbuses. Currently, Ethernet is increasingly used in DCCS, however, the available solutions to guarantee real-time behaviour in Ethernet are not well suited to support the DCCS requirements and can't be implemented in low processor power devices. This paper proposes a real-time Ethernet-based protocol (VTPE) that supports real-time requirements, and implies just a small overhead in the system nodes. The proposed protocol is based on the virtual token-passing media arbitration and does not require specialized hardware.

This paper considers the case in which master-slave fieldbus networks are used in safety-critical embedded applications, such as transportation systems. Traditional approaches to system design, due to fault-tolerance reasons, have considered static cyclic table-based traffic scheduling, only. However, there is a growing demand for flexibility and integration, mainly to improve efficiency in the use of system resources, with the network playing a central role to support such properties. This calls for dynamic on-line traffic scheduling techniques so that dynamic communication requirements are adequately supported. This paper considers such dynamic master-slave architectures and addresses the problem of enforcing fail silent behavior both in the master and in the slave nodes. Two different mechanisms are proposed, one based on dynamic bus guardians for the slave nodes only, to impose fail silent behavior in the time domain, and other based on internal replication and temporized agreement, to impose fail silence both in the temporal and value domains. Despite being potentially applicable to a set of master-slave networks, this paper discusses the specific implementation of the proposed mechanisms on top of the FTT-CAN protocol.

A flexible distributed real-time communication system must support modifications to the message set which it conveys. These changes can require a degree of responsiveness ranging from a few milliseconds to some seconds. In FTT-CAN protocol, the responsiveness of the synchronous communication system depends on the plan duration, which, in general cannot be set arbitrarily short. This paper presents a method that uses the asynchronous messaging system to temporarily convey the synchronous messages until the synchronous messaging system can handle them. Furthermore, methods to evaluate offline if a set of requests for modifications can be timely handled are presented.

Distributed control systems performance can be affected by the occurrence of jitter in the messages that carry relevant data such as sample and actuation variables. This jitter comes from the influence of messages from other sources and thus depends on factors such as the distribution of controller tasks and the medium access control used in the network. When a CAN- controller area network is considered, this jitter can be modelled as a random variable with a gamma distribution. In this paper a study of the influence of this specific type of jitter in system identification with recursive implementation is presented. The results are derived in an adverse situation when both sampling and actuation data suffer from jitter. It is shown that using a model that assumes fractional dead-time in the system leads to a much better parameter identification than when the problem is just ignored.

The requirement for flexible operation is becoming increasingly important in many distributed computer control systems. This paper briefly presents the FTT-CAN protocol (Flexible Time-Triggered communication on Controller Area Network) which allows for flexible operation, i.e. dynamic communication requirements, under timeliness guarantees. Then, the paper discusses a particular approach to the global management of such distributed systems named network-centric. This is based on the synchronisation of the distributed tasks in remote nodes with the network traffic. Thus, on-line changes in the rates and/or relative phasing of the network traffic can be automatically applied to the tasks that use it, in all the remote nodes in the distributed system. This approach is fully supported by the FTT-CAN protocol.

The operational flexibility of fieldbuses is becoming an important issue in modern industrial automation systems. In these systems, control and measurement equipment is normally connected to a fieldbus. For certain applications it is interesting to dynamically change the system configuration by either removing or adding equipment, or just by changing certain properties of existing one. The difficulty in doing so is to guarantee that the modifications done on-line do not cause any timing constraints in the communication system to be missed. This article discusses the existing approaches to real-time message scheduling in what concerns the schedulability analysis of bus traffic. The use of a planning scheduler, as proposed previously by the authors, is briefly described and the impact on system flexibility, scheduler run-time overhead and schedulability guarantee is discussed. Some new results concerning the worst-case execution time of the scheduler and the worst-case message delays are presented. The article also includes a discussion on the applicability of the planning scheduler in the context of existing fieldbus standards. It is shown that this scheduler can improve the operational flexibility of fieldbuses based on static scheduling approaches and allow on-line schedulability analysis in fieldbuses based on dynamic best-effort approaches.

A typical approach to real-time fieldbus arbitration is the use of an off-line scheduler that generates a cyclic static table containing the allocation of bus-time-slots to the transaction of process control variables. This approach, e. g. used in the FIP fieldbus (Factory Instrumentation Protocol), is highly inflexible in the sense that any system changes, such as adding a sensor, requires the interruption of the fieldbus operation. In this article the use of a planning scheduler is proposed to overcome such inflexibility. This scheduler compromises between the advantages and disadvantages of typical dynamic and static scheduling. The possible use of the planning scheduler within the FIP context is also described resulting in a compatible FIP-like fieldbus.

The theory of real-time task scheduling can be applied, in many situations, to real-time message scheduling. Particularly, it can be advantageously applied in real-time fieldbus networks such as the FIP fieldbus (Factory Instrumentation Protocol). In this paper the Rate-Monotonic Scheduling theory is used to derive a suficient schedulability condition for the FIP fieldbus with a modified scheduler. Such condition may enhance the usability of FIP-like fieldbuses in dynamic environments.