Contact Information

Networked Real-Time RFID System

Motivation

The real-time availability of RFID data is critical for many RFID applications, such as manufacturing automation systems, process control systems, advanced materials handling systems and supply chains. For example, in a manufacturing system described in Figure 1, the materials control system (MCS), which controls the flow of physical items, wants to be updated of the identity and location of a work-in-process (WIP) in real-time. The process control system needs to have the sensor data in real-time, make control decisions in real-time and execute the decisions in real-time. While the RFID technology could provide important identity and location information of items, it is predictable that RFID technology will eventually be integrated into those systems. Therefore, we feel it is very important to research the real-time operability of networked RFID systems for those applications.

Figure 1: Real-Time RFID Manufacturing System

Definition

Real-time systems are information processing systems which have controls over on external environments, and have to respond to stimuli generated by external environments and affect the functioning of the external environments within a finite and specified period. The information processing time constraints are derived from the dynamics of the physical objects or processes under control in the external environment s . The correctness of the system s depends not only on the logical results of the computation but also on the time when the results are generated.

A networked RFID system generally comprises the following elements:

1. A unique identification number which is assigned to a particular item.

2. An identity tag that is attached to the item with a chip capable of storing - at a minimum - the unique identification number. The tag is capable of communicating this number electronically.

3. Networked RFID readers and data processing systems that are capable of collecting signals from multiple tags at high speed (100s per second) and of pre-processing this data in order to eliminate duplications, redundancies and misreads.

4. One or more networked databases that store the product information.

With this approach, the cost of installing and maintaining such systems can be spread across several organizations while each is able to extract its own specific benefits from having uniquely identified items moving in, through and out of the organization’s operations.

Research Problems

When RFID technology is integrated to real-time control systems, to ensure the real-time operability of the system, that is, the system must meet any timing constraints set by the environment, most commonly, the timing constraint is that the control information flow must be quicker than physical flow, we need to know:

• What is the system timing constraint set by the environment?
• How could the system meet the timing constraint?

The first question is easy to answer; it is the timing constraint or deadline set by the controlled process or physical system for the information process system, which is rather explicit. Deadlines in real-time systems are usually introduced to specify quality of service (QoS) or control the operation of physical systems.

The second question is not so explicit, but it could be answered by decomposing the problem into the following sub problems:

• How could the system be decomposed into different subsystems or processes?

The system could be decomposed into the following functional blocks, as shown in Figure 2. Each functional block could then be decomposed into time delay blocks in Figure 3.




>
Figure 2: RFID control system functional blocks diagram		Figure 3: RFID Control System time delay blocks diagram




• What is the timing performance of each subsystem?
• How could the system timing constraint be divided into constraints for each subsystem or process?
• What is the performance of each subsystem in terms of meeting respective timing constraints?
  Probabilistic description of the performance is desirable. In a hard real-time system, an operation is considered successful only if it finishes its execution within its specified deadline. Therefore, the key measure of system performance is the percentage of all operations that complete within their deadlines.
• In terms of timing performance of each subsystem or process, how to schedule all the processes in order to design such a system that could meet the timing constraint?
  Concurrency control protocols need to be considered together with scheduling algorithms.
• In terms of design, how to specify the timing constraints of system behaviour and verify the timing constraints have been met?
  It will need to apply formal methods for specification and verification of real-time systems.

The first sub-question is easy to answer by following the block diagram rules of control systems. The second sub-question could be answered by experiments results or practical data. The third sub-question depends on the answer to the second sub-question. It could also be done in an iterative way by returning to this sub-question after answering the fourth sub-question. The fifth sub-question is the most important issue in real-time systems, in which scheduling mechanism and concurrency control protocols should be applied. It could be answered by comparing different scheduling algorithms and concurrency control protocols, with regard to the performance of meeting timing constraints. The sixth sub-question is about the timed systems modeling and models checking, it could be answered by comparing different modeling approaches.

Research Aims

Integration of Radio Frequency Identification and Electronic Product Code network and real-time operations, evaluating the current real time limitations using our lab as an example and then determine how many of these limitations can be overcome for high speed requirements.

Figure 4 shows a real-time RFID experiment which could be done in Automation Lab. Figure 5 is the time delay diagram of the experiment.

Learn more..

For more information about this project, please contact Wei Wang William Wong [at] ieee.org