A year ago, a CIO.com headline proclaimed 2022 as “the Year of the Edge”. Now, several others are proclaiming 2023 “the Year of Edge AI” and several forecasters expect the whole Edge AI market value to sixfold before 2030. The accelerating demand for resource-efficient computing and green AI can take it even higher.
Greentech, medtech, biotech, miltech, aerotech, fintech, Industry4.0 tech, you-name-it tech; the synergies of Edge computing and AI/ML improve real-time quality parameters (performance, availability, adaptability in runtime and fast dev-ops, data ownership and integrity/PII/GDPR) as well as the cost-benefit ratio (reduced power consumption, network footprint, HW cost, to name a handful). Edge ML takes the algorithm close to the devices that generate the training-data sets, and uses cheap resource-saving local HW in training time as well as in runtime/inferencing.
Decentralized ML architectures are kins of Edge patterns (three basic Edge architecture patterns: our Agile Architecture course). So, let’s keep edging in AI, for a rapid RT response to new risks or opportunities, plus for many other environmental and financial reasons. The complexity and computing power of Edge devices ranges from relatively modest (for example, simple wearables for patients during tests of new vaccines) to very high in aerospace where an aircraft operating under strict radio silence is a huge “Edge device” with an onboard network of smaller Edge devices, typically interconnected by an onboard event bus.
According to Mats Palmberg of Saab, Gripen E’s state-of-the-art electronic warfare suite, with 360 degree spheric protection, beats stealth; radars detect targets that are ten times smaller than usual. Passive sensors make sure the position of the aircraft is not compromised. Gripen can be integrated with jammers and decoy missiles for more demanding missions.
Today’s aircraft, especially Gripen with its flexibility and record-low CPFH (Cost per flight hour) of appx 31,000 SEK, makes a textbook case of the benefits of Edge computing for any IT architect in itssector of industry– and in many other sectors too.
Trainer at Informator, senior modeling and architecture consultant at Kiseldalen’s, main author: UML Extra Light (Cambridge University Press) and Growing Modular (Springer). Advanced UML2 Professional (OCUP cert level 3/3). Milan and Informator collaborate since 1996 on architecture, AI, rules, modeling, UML, requirements, and design. You can meet him this year at these courses in English or Swedish (remote participation is offered and recommended) :
AI, Architecture, and Machine Learning
Agile Architechture Fundamentals
Avancerad objektmodellering med UML
(on demand: Modular Product Line Architecture