Control system design is a crucial point in modern automation systems, involving an increasingly broader set of activities being performed by software. Despite the availability of several domain-specific facilities, designers daily face kindred issues which actually lack reference patterns, in particular as regards the architectural organization of the control logic. In this paper we present an approach that, focusing on cross-cutting domain problems and suitably abstracting from application-specific details, leads to an effective decoupling of high-level control policies from low-level actuation and sensing mechanisms. Logic Device Agents emerge as generally applicable active components modeling the behavior of commonly used field devices, which can be profitably exploited to enhance software design quality and productivity.

Most organizations show a strong interest in digital signature technology as a means for secure and authenticated document exchange, hoping that it helps reduce the paper-based transactions. The main problem posed by this technology is with the necessary public-key infrastructure, and in particular with certificate status handling. Rather than addressing the revocation problem in general, a specific but interesting aspect is discussed here: secure identification of a large number of users (like citizens for a public administration) accessing a wide pool of services. This paper describes the definition and deployment of a web-based environment suitable for offering administrative services to citizens and for accepting authenticated documents from citizens. The best features of two different certificate status handling schemes, the standard CRL and a novel on-line scheme, have been exploited within this environment to obtain a good balance between security, timeliness and efficiency.

A new on-line method for efficient handling of certificates within public-key infrastructures (PKIs) is presented. The method is based on a purposely-conceived extension of the one-way accumulator (OWA) cryptographic primitive, which permits one to provide an explicit, concise, authenticated and not forgeable information about the revocation status of each certificate. A thorough investigation on the performance attainable shows that the devised method exhibits the same positive features of the well-known on-line certificate status protocol (OCSP) as regards scalability, security and timeliness. Moreover, its peculiar characteristic of collectively authentication via a single directory-signed proof the status of all the certificates handled within a PKI leads to a significant reduction of the directory computational load that, in a high-traffic context, could be nearly unbearable when OCSP is applied.

The widespread use of public networks, such as the Internet, for the exchange of sensitive data, like legally valid documents and business transactions, poses severe security constraints. The approach relying on public-key certificates certainly represents a valuable solution from the viewpoint of data integrity and authentication. The effectiveness of the approach, however, may be arguable, especially when a trivial strategy is adopted within a public key infrastructure (PKI) to deal with the problem of revoked certificates. This paper presents a novel certificate status handling scheme, based on a purposely-conceived extension of the one-way accumulator (OWA) cryptographic primitive. The distinguishing characteristic of the devised Owa-based Revocation Scheme (ORS) is that it exploits a single directory-signed proof to collectively authenticate the status of all the certificates handled by a certification authority (CA) within a PKI. A thorough investigation on the performance attainable shows that ORS exhibits the same features of the well-known Online Certificate Status Protocol (OCSP) as regards security, scalability and certificate status-updating timeliness, at the same time drastically reducing the directory computational load that, in a high-traffic context, could be nearly unbearable when OCSP is applied.

Addresses the grasp synthesis problem arising in task planning for robotic dextrous hands. For this purpose, a hybrid architecture is proposed, which relies on symbolic and subsymbolic computations to exploit heterogeneous sources of knowledge, such as practical experience learned through experiments with a real device, heuristic rules gained from human observation, geometric reasoning and, when applicable, analytical results. After a preliminary discussion of representation levels and techniques, this paper describes the design of a tool for selecting the feasible grasps of a robotic hand under various situations and for ranking them according to task-oriented criteria. The interaction of a rule-based expert system with a neural network-based classifier provides support to both explicit reasoning and direct learning from experience. The features taken into account by the tool are object geometry, hand kinematic capabilities, task requirements, (in terms of both accessibility and robustness) and workspace constraints.