{"id":27000,"date":"2026-09-27T20:02:20","date_gmt":"2026-09-27T18:02:20","guid":{"rendered":"https:\/\/www.itspy.cz\/thesis\/satellite-telemetry-anomaly-detection\/"},"modified":"2026-10-23T13:46:12","modified_gmt":"2026-10-23T11:46:12","slug":"satellite-telemetry-anomaly-detection","status":"publish","type":"thesis","link":"https:\/\/www.itspy.cz\/en\/thesis\/satellite-telemetry-anomaly-detection\/","title":{"rendered":"Satellite telemetry anomaly detection"},"content":{"rendered":"<p>This thesis addresses anomaly detection in satellite<br \/>\ntelemetry, where anomalies may signal impending system<br \/>\nfailures with potentially catastrophic consequences. The<br \/>\nwork investigates the use of artificial intelligence<br \/>\nmethods for this task. Multiple detection models were<br \/>\nimplemented and evaluated, including long short-term memory<br \/>\n(LSTM) networks, autoencoders, residual networks (ResNet),<br \/>\nand convolutional neural networks, using both<br \/>\nforecasting-based and reconstruction-based approaches.<br \/>\nModel performance was assessed using several metrics, with<br \/>\nPR-AUC (area under the precision\u2013recall curve) as the<br \/>\nprimary criterion. The experiments were conducted on a<br \/>\nsynthetic telemetry dataset generated by a custom data<br \/>\ngenerator developed as part of this work. The best results<br \/>\nwere achieved by an LSTM-based autoencoder model, reaching<br \/>\na PR AUC of 0.946. To evaluate deployment feasibility,<br \/>\ninference performance was tested on embedded platforms (AMD<br \/>\nZCU104 and NVIDIA Jetson Orin Nano), simulating onboard<br \/>\nsatellite computing constraints. In addition, real<br \/>\ntelemetry data from the ESA anomaly detection benchmark<br \/>\nwere used to assess performance in practical scenarios.<br \/>\nFinally, an Anomaly Exploration and Report System (AERS)<br \/>\nwas developed to complete the detection pipeline, providing<br \/>\nan interface for domain experts to implement custom anomaly<br \/>\nexplanation modules.<\/p>\n","protected":false},"featured_media":26763,"template":"","meta":{"_acf_changed":false,"_links_to":"","_links_to_target":""},"university":[189,138,441],"thesis-year":[466],"class_list":["post-27000","thesis","type-thesis","status-publish","has-post-thumbnail","hentry","thesis-year-466"],"acf":{"autor":"Vojt\u011bch Orava","portret":"","vedouci":"Ing. Tom\u00e1\u0161 Ka\u0161p\u00e1rek, Ph.D."},"_links":{"self":[{"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/thesis\/27000","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/thesis"}],"about":[{"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/types\/thesis"}],"version-history":[{"count":1,"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/thesis\/27000\/revisions"}],"predecessor-version":[{"id":27055,"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/thesis\/27000\/revisions\/27055"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/media\/26763"}],"wp:attachment":[{"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/media?parent=27000"}],"wp:term":[{"taxonomy":"university","embeddable":true,"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/university?post=27000"},{"taxonomy":"thesis-year","embeddable":true,"href":"https:\/\/www.itspy.cz\/en\/wp-json\/wp\/v2\/thesis-year?post=27000"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}