Future of Engineering
Download as PPTX, PDF2 likes8,983 views
The document discusses the future of engineering and how various technologies will impact the field. It predicts that engineering will become more multi-disciplinary, involve less human workers and more automation through tools like AI and machine learning. Specific technologies like GPUs and neural networks will allow extremely powerful yet small supercomputers to be used for engineering tasks. Overall, the boundaries between software, machine learning and different engineering disciplines will blur as these fields converge and new approaches like "learning not programming" become more common.
Future of Engineering
- 1. Future of Engineering Anatoly Levenchuk, TechInvestLab president, INCOSE Russian chapter director for research ailev@asmp.msk.su Moscow 20-Jul-2017
- 2. 2 The future is already here — it's just not very evenly distributed. William Gibson http://www.johnsonsdrivingacademy.co.uk/driving-advice/driving-in-the-fog/ Geoffrey Hinton metaphor: fog is completely transparent up to 5 years, sharply non-transparent further.
- 3. Engineering is not digital yet 3 CB Insights https://twitter.com/CBinsights/status/705403757916528640 3 March 2016
- 4. Everything will be fast http://blogs.nvidia.com/blog/2016/01/12/accelerating-ai-artificial-intelligence-gpus/ By 2011 AI researchers around the world had discovered NVIDIA GPUs. 2012 – first victory of ImageNet competition 2015 – superhuman performance 2017 – last ImageNet competition
- 5. 5http://www.nvidia.com/object/embedded-systems-dev-kits-modules.html NVIDIA Jetson TX-2 – 7.5Watt, 1TFLOPS supercomputer, 50*87mm Exponential: about twice cheaper yearly (Moore law today: GPU, not CPU) delivering with software AI will be is a commodity
- 6. Intelligence Platform Stack 6 Application (domain) Platform engineering Cognitive Architecture Platform Learning Algorithm Platform Computational library General Computer Language CPU GPU/ TPU/DPU/FPGA/Physical computation Drivers GPU/ TPU/DPU/FPGA/Physical computation Devices Neurocompiler Neuromorphic driver Neuromorphic chip Disruptionenablers Disruptiondemand Thanks for computer gamers for their disruption demand to give us disruption enabler such as GPU!
- 7. Change of the Life Cycle Models 7 whatchmaker forestergardener Planning and designing in engineering Growing and learning in engineering
- 8. Different kinds of engineering merging and disruprtion 8 Software Engineering Machine Learning Engineering+
- 9. Future of engineering1. Extremal – Multi-scale – Multi-disciplinary – Starts before traditional life cycle – Ends after traditional life cycle – More First-in-a-Kind systems – Reverse engineering (including for verification/compliance purposes) – Using system (not only system of interest) re-engineering 2. Less people, more computers (automatization) – Generation (beyond model-based systems engineering) – Learning, not programming – Avatarization of software – Alexa/Siri/Cortana engineering skills – Virtual reality for industrial design purposes – Software as a service, clouds + edge computing (CAD/PLM in a system for resilience) 3. Multiagent engineering – Cyborg-2 (agents: people and machines) – Disintermediation of engineers (no needs for companies) – Open engineering (open architecture first of all) – Advanced modularity (accumulation of knowledge) 4. All life cycle stages (system definition AND realization AND operation) 9
- 10. Innovator’s Dilemma: its true for engineering too (Clayton Christensen from Harvard University) c h a r a c t e r i s t i c s time The best products on the market The worst products on the market http://web.mit.edu/6.933/www/Fall2000/teradyne/clay.html 10
- 11. 11 Спасибо за внимание Anatoly Levenchuk, http://ailev.ru ailev@asmp.msk.su 20 June 2017