Spring 2020 Project Presentations *********************************************** Note: Each presentation is around 15 minutes. For a team project, each member of the team must present. Everyone is expected to attend and ask questions (attendance will be taken). A discussion follows each talk. If you are at the beginning or midway through your project, present an outline of what you are doing and expected results. Write a short review (a paragraph) of each talk for submission with your project report (in pdf). Due: Monday 5:00pm, 4/27/2020. Extended to 5:00pm, 5/4/2020. Also send me the presentation slides after the talk. Please email me your topic and referenced paper(s)' titles at least one week before your presentation. *********************************************** 4/20 Garcia,Patricio Topic: Implementing Self-Stabilizing Techniques For Real-Time Expert Systems With Bounded Response Time in C References: A. M. K. Cheng, “Self-Stabilizing Real-Time Rule-Based Systems,” https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/document/235129 A. M. K. Cheng and S. Fujii, “Self-stabilizing real-time OPS5 production systems” https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/document/1350764 Jeng-Rung Chen and A. M. K. Cheng, “Predicting the response time of OPS5-style production systems” https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/document/378822 B. Zupan and A. M. K. Cheng, “Optimization of rule-based systems using state space graphs” https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/document/683755 A. M. K. Cheng, Real-Time Systems: Scheduling, Analysis, and Verification, Chapter 10 ISBN # 0-471-18406-3 --------------------------------------------------------------------------- Grimes,Zachary Topic: Implementing machine learning and sensors that interface with a board such as Beaglebone AI / Jetson Nano. Camera sensor analysis (real time prediction accuracy and computation times, power consumption, etc.). --------------------------------------------------------------------------- Neugebauer,Michael Thomas --------------------------------------------------------------------------- Nguyen,Frank P Topic: Analysis of PO-WFD and PAPO-WFD in multiprocessor systems. Performance analysis of these proposed algorithms (overhead to map using the algorithms mathematically) and their implementation (if time allows). References: An efficient method for assigning Harmonic periods to Hard Real-Time Tasks with Period ranges - https://ieeexplore.ieee.org/document/7176034 Preference-Oriented Scheduling in Multiprocessor Real-Time Systems - https://ieeexplore.ieee.org/document/8603199 Preference-Oriented real-time scheduling and its application in fault-tolerant systems (Not on IEEE, but pdf is somewhere on the web) --------------------------------------------------------------------------- 4/22 Romero,Pallovi Felipa Topic: Multiprocessor Earliest Deadline First scheduling of Industrial Robotic Manufacturing tasks References: Pallovi Romero and Albert M. K. Cheng, ``EDF Scheduling of Industrial Robotic Manufacturing Tasks,'' International Symposium on Measurement, Control, and Robotics (ISMCR), University of Houston-Clear Lake, Texas, USA, September 19-21, 2019. Albert M. K. Cheng, Real-Time Systems: Scheduling, Analysis, and Verification (John Wiley & Sons) ISBN # 0-471-18406-3, 552 pages, August 2002; 2nd printing with updates, 2005. --------------------------------------------------------------------------- Louie,Andrew C Topic: Analyzing and optimizing Compression of Images for Real-Time Applications using anytime image segmentation, super pixels and regional compression. Priority-Driven Coding of Progressive JPEG Images for Transmission in Real-Time Applications, Albert Mo Kim Cheng and Feng Shang https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/stamp/stamp.jsp?tp=&arnumber=1541069 Real-Time Superpixel Segmentation by DBSCAN Clustering Algorithm, Jianbing Shen, Xiaopeng Hao, Zhiyuan Liang, Yu Liu, Wenguan Wang, and Ling Shao https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/document/7588062 Scheduling of Image Processing Using Anytime Algorithm for Real-time System, Wyne Wyne Kywe, Daisuke Fujiwara, and Kazuhito Murakami https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/stamp/stamp.jsp?tp=&arnumber=1699716 Real-time Image Processing based on a Task-pair Scheduling Concept, Thomas Ihme , Kai Wetzelsberger , Mario Speckert , Jorn Fischer https://ieeexplore-ieee-org.ezproxy.lib.uh.edu/stamp/stamp.jsp?tp=&arnumber=5980172 Compound Image Compression for Real-Time Computer Screen Image Transmission, Tony Lin and Pengwei Hao --------------------------------------------------------------------------- 4/27 Sun,Zhiyuan (Did not present.) --------------------------------------------------------------------------- Torre,Elena Virginia Topic: Fault-Tolerant Real-Time Virtualization References: Albert M. K. Cheng, Guangli Dai, Pavan Kumar Paluri, Mansoor Ansari, Yu Li and Darrell Brandon Knape, “Fault-Tolerant Regularity-Based Real-Time Virtual Resources,” 25th IEEE International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), Hangzhou, China, August 18-21, 2019. Yu Li and Albert M. K. Cheng, “Toward a Practical Regularity-based Model: The Impact of Evenly-Distributed Temporal Resource Partitions,” ACM Transactions on Embedded Computing Systems (TECS), Volume 16, Issue 4, Article No. 111, August 2017. Yu Li and Albert M. K. Cheng, “Transparent Real-Time Task Scheduling on Temporal Resource Partitions,” IEEE Transactions on Computers (TC), Volume 65, Issue 5, pages 1646-1655, May 2016. Yu Li and Albert M. K. Cheng, “Static Approximation Algorithms for Regularity-based Resource Partitioning,” Proc. 33rd IEEE Real-Time Systems Symposium (RTSS), San Juan, Puerto Rico, USA, December 4-7, 2012. --------------------------------------------------------------------------- Wert,Jonathon Russell Topic: Improving the inference response time of a machine learning (ML) model Description: My idea was to use something simple like the iris data set. I could train a model and wrap it inside a simple Python API (flask) and containerize it with Docker. + Using a raspberry pi 3, as a local machine and simulated embedded device, I could test local queries and track response times with no accelerated hardware. + Then I could deploy the Docker to the cloud and test the query response times on the raspberry pi simulating an embedded device that utilizes cloud resources for inference. + Then I could accelerate the raspberry pi 3's inference ability by attaching Google's usb accelerator called Coral, which features a Google Edge TPU and test response times with acceleration, simulating an embedded device without need for network connection. I could make comparisons and cost effectiveness of deploying to the cloud versus running inferences on locally accelerated hardware. The benefits of analyzing data in the field where network connections may not be available, especially at speeds required for real time feedback. --------------------------------------------------------------------------- Rodriguez,Ivan A Topic: Determining WCETs for a Hewlett Packard Enterprise software component. References: https://www.hpe.com/us/en/integrated-systems/software.html --------------------------------------------------------------------------- Rezaei Firuzkuhi,Mahsa (Did not present.) Topic: “An Implementation of Elastic Scheduling for Real-Time Tasks” [1] G. C. Buttazzo, G. Lipari, M. Caccamo and L. Abeni, "Elastic scheduling for flexible workload management," in IEEE Transactions on Computers, vol. 51, no. 3, pp. 289-302, March 2002. [2] T. Chantem, X. S. Hu and M. D. Lemmon, "Generalized Elastic Scheduling," 2006 27th IEEE International Real-Time Systems Symposium (RTSS'06), Rio de Janeiro, 2006, pp. 236-245. [3] T. Chantem, X. S. Hu and M. D. Lemmon, "Generalized Elastic Scheduling for Real-Time Tasks," in IEEE Transactions on Computers, vol. 58, no. 4, pp. 480-495, April 2009. [4] James Orr, Chris Gill, Kunal Agrawal, Jing Li, and Sanjoy Baruah. 2019. Elastic Scheduling for Parallel Real-Time Systems. Leibniz Transactions on Embedded Systems 6, 1 (2019). [5] James Orr and Sanjoy Baruah. 2019. Multiprocessor scheduling of elastic tasks. In Proceedings of the 27th International Conference on Real-Time Networks and Systems (RTNS ’19). Association for Computing Machinery, New York, NY, USA, 133–142. ---------------------------------------------------------------------------