RETIS Lab Real-Time Systems Laboratory

Transcription

1 RETIS Lab Real-Time Systems Laboratory A brief introduction RETIS Lab Real-Time Systems Laboratory RETIS Lab investigates embedded computing systems with particular emphasis in Real-time methodologies Operating systems Design tools 1

2 People RETIS Lab Real-Time Systems Laboratory Currently, the RETIS Lab comprises 35 people, including: 2 Full professors 2 Associate professors 4 Assistant professors 12 Research associates 15 PhD students What is an embedded system? RETIS Lab Real-Time Systems Laboratory It is a computing systems hidden in an object to control its functions, enhance its performance, manage the available resources and simplify the interaction with the user. Object micro- processor actuators sensors Environment communication user other units 2

3 Real-Time Systems Laboratory Embedded computing systems are becoming pervasive in our society: Flight control systems Plant control Automotive RETIS Lab Railways switching systems Robotics Defense systems Patient monitoring systems and many others 3

4 What s special in Embedded Systems? FEATURES Scarce resources (space, weight, time, memory, energy) High concurrency and resource sharing (high task interference) Tight environment interaction High variability on workload and resource demand REQUIREMENTS High efficiency in resource management Limit interference by temporal isolation High predictability in the response time Robustness (Overload handling and adaptivity) Predictability + efficiency The RETIS group has contributed to increase predictability and efficiency of real-time systems for flight control and automotive applications. In collaboration with Airbus and Bosch Airbus Bosch 4

5 Predictability in Linux Linux has been extended to handle real-time software and support the rapid growth of complexity in next-generation cell phones. In collaboration with Ericsson ARM11 (4 cores) Linux 2.6.x The FLEX Board In collaboration with Microchip 5

6 Embedded Control Board Main features: Power management Remote programming Expandibility: Connection with other special-purpose boards Architecture Software Application (C) RTOS Hardware ERIKA Microchip dspic 30F601x 6

7 Educational control kits Objectives ASCOLTA Home assistance for heart failure through an advanced wearable monitoring systems Reduce hospital costs through continuous monitoring at home. Simplify health monitoring through a wireless wearable monitoring system Automatic detection of critical conditions, asking for medical intervention. 7

8 ASCOLTA SPO2 ECG Blood pressure Breath signal Body motion Wireless Sensor Networks actuators Controller HW sensors Environment Node wireless Typical measures sound light gas temperature video Scuola Superiore Sant Anna 16 8

9 Mobile Nodes Scuola Superiore Sant Anna 17 Sample applications Surveillance Intrusion detection and tracking Defence systems Environmental monitoring Exploration Rescuing 18 9

10 Fire prevention... Scuola Superiore Sant Anna 19 Traffic monitoring Scuola Superiore Sant Anna 20 10

11 International Collaborations Carnegie Mellon University Prof. Rajkumar / Lehoczky (feasibility analysis) Univ. of Illinois at Urbana Champaign Prof. Lui Sha / Marco Caccamo (RT control and scheduling) Univ. of North Carolina at Chapel Hill Prof. Sanjoy Baruah (Multiprocessor scheduling) Univ. of California at Berkeley Prof. Sangiovanni Vincentelli (Design methodologies and tools) Florida State University Prof. Ted Baker (Multiprocessor scheduling) Univ. of Indianapolis Prof. Yao Liang (Sensor Networks) RETIS Lab Real-Time Systems Laboratory Trondheim Vasteras Goteborg Stockholm Cork York Lund contacts Eindhoven Dresden Dortmund Prague Rennes Nancy Kaiserslautern Wien Zurich Porto Cantabria Madrid Grenoble Pv Fi Si Tr Bo An Barcelona Valencia Catania 11

12 RETIS Lab Real-Time Systems Laboratory UIUC Philadelphia Pittsburgh Berkeley Virginia UNC Austin FSU Industrial Collaborations Philips - QoS management for multimedia systems Ericsson RT t f t l i ti - RT systems for telecommunications Airbus Bosch Microchip Magneti Marelli Ansaldo - kernels for avionic applications - kernels for automotive applications - RTOS for embedded control applications - Schedulability tools for automotive systems - Predictable kernel mechanisms ST Microelectronics - RTOS for multicore architectures 12

13 Courses on Embedded Systems Course Teacher NOTES Real-Time Systems Introduction to Neural Networks Object Oriented Software Design in C/C++ Giorgio Buttazzo Giorgio Buttazzo Giuseppe Lipari Methodologies for increasing software predictability. Task scheduling and feasibility analysis; protocols for accessing shared resources; interrupt handling. Basic networks models and learning algorithms for pattern recognition, classification, prediction, and control. Software design methodologies in C++; Code design patterns; Templates; Concurrency in C++; the new C++11 standard. Courses on Embedded Systems Course Teacher NOTES Embedded Systems Marco Di Natale Introduction to the development of programs and models for embedded software. Finite state machines. Automatic code generation using Simulink tool from Mathworks. Laboratory of Real-Time Systems Mauro Marinoni, Gianluca Franchino Programming microcontrollers, I/O devices, sensor acquisition and processing, servomotor control, wireless communication protocol. Use of a real-time kernel to develop real-time control applications. 13

14 Courses on Embedded Systems Course Teacher NOTES Combinatorial Optimization Android Framework Dependability and Fault-Tolerance Michela Milano Andrea Roli Enrico Malaguti Michele Lombardi Alberto Panizzo Felicita Di Giandomenico Integer Linear Programming Constraint Programming Simulated Annealing Genetic Programming Hybrid methods Mobile device features, Application environment, User interface, Data management Complex services and security Fault models. Error detection. Forward and backward error recovery. Static and dynamic redundancy. Transactions and checkpoints. Courses on Embedded Systems Course Teacher NOTES Modeling of concurrent programming Sverre Hendseth Formal methods for expressing concurrent programs and verifying their correctness. How to do research Giorgio Buttazzo Gerhard Fohler How to write scientific papers Common English mistakes How to make slides How to make presentations How to review papers 14