took place virtually in Vienna on 2-3 September.<\/span><\/p>\nToday formal methods – techniques more than 50 years old – is recognized by the FMICS community as cutting-edge technologies for specification, development, and verification of software and hardware systems.<\/span><\/p>\nFMICS 2020 was an unprecedented effort to collect knowledge of formal methods, based on an extensive survey of 130 internationally renowned experts. Based on their answers and comments were proposed a rigorous analysis of the past, present, and future of formal methods in research, education, and industry was presented.<\/span><\/p>\nAccording to a vast majority of the experts (almost 100%), formal methods are very important for building safety-critical software. They contribute to the construction of trustworthy systems and are crucial for high-quality assessments. Moreover, 73,9 % of respondents estimated that formal methods lead to better performance, 60% – lower cost, and 56,9 % faster development.<\/span><\/p>\nFormal methods are very important for safety-critical software in the railway domain. The CENELEC EN 50128 standard recommends formal methods for the highest safety integrity level (SIL).<\/span><\/p>\nStefan Resch (Thales Austria GmbH) \u2013 one of the invited speakers \u2013 highlighted successful applications of formal methods in the railway industry:<\/span><\/p>\n1)<\/span>\u00a0 <\/span> ERTMS Hybrid Level 3 (successful use of FMs for increasing track capacity in the railway network);<\/span><\/p>\n2)<\/span>\u00a0 <\/span> Checking European Train Control System (ETCS) Level 1 Lineside configuration data (many data errors can be caught in the early stages by using the Emerald approach);<\/span><\/p>\n3)<\/span>\u00a0 <\/span> TAS Control Platform (analyzing core safety and liveness properties of a fault-tolerant protocol by using model-checker).<\/span><\/p>\nA huge majority of the experts in the survey also agreed about the unique qualities of formal methods in providing correctness, which makes it incomparable to other methods such as artificial intelligence, quantum or molecular computing.<\/span><\/p>\nRegardless of the technical benefits of formal methods, there are still several obstacles to making them more spread in the industry. The most important of them originate in financial and human factors.<\/span><\/p>\nAround 60% of experts thought that the application of formal methods was profitable only in the medium and long term. Immediate return is prevented by the high initial investment cost: \u201cThe real cost savings appear only later, with less and cheaper maintenance due to fewer failures\u201d.<\/span><\/p>\nThe human factor is another potential barrier to the wider usage of formal methods in the industry. Formal methods depend on sufficiently high skill levels and can be difficult to learn and deploy. According to the experts’ opinions, both managers and developers show a lack of knowledge in formal methods: \u201cDevelopers are reluctant to change their way of working\u201d.\u00a0 Many experts also emphasized the lack of appropriate scientific\/technical education. Expert Joost-Pieter Katoen underlined in his statement the need for “lightweight” formal methods.<\/span><\/p>\nThe application of formal methods is foreseen by many experts to spread in the future, not only to be applied in software engineering and computer science (69,2 %), but also in other science (biology, chemistry, epidemiology, etc. \u2013\u00a0 58,5 %), finance (digital currencies, smart contracts, etc. \u2013 61,5%), in politics (e-voting, e-government, etc. \u2013 43,8 %) and other parts of society \u2013 31,5%.<\/span><\/p>\nEinar Broch Johnson claimed in his statement that formal methods can be applied \u201canywhere in science or society, where correct models matter\u201d.\u00a0 <\/span><\/p>\n<\/div><\/div><\/div><\/div><\/div>![](\"https:\/\/www.prover.com\/wp-content\/uploads\/2024\/07\/safety.png\" "\\"safety\\"")
<\/span><\/div><\/div><\/div>