3/11/2017 Decision making in complex systems Workshop Facilitators: Dr Mal Tutty, Dr Keith Joiner, Luke Brown
Workshop Abstract Over the last three decades, defence communication and information systems have been increasing the complexity and interconnectedness of systems that has pervaded society more broadly throughout the Information Age. Even more than society in the broad, Western Departments of Defence (DoDs) have sought to attain information dominance. The result has been a large number of complex systems, system-of-systems and families-of-system-of-systems. This workshop examines the Australian implications of the big four challenges facing decision making in complex systems and the six key assurance initiatives pursued systematically by the US DoD initiatives to effect these more integrated, interoperable and information-assured (I3A) capabilities, while also ensuring these capabilities remain resilient to the new cyber threats.
Challenges Challenge 1 - Growing System Synthesis Challenge 2 Higher-Order Human Functions Challenge 3 Cyber Threat Complexity response to Information Dominance Challenge 4 - Requirements Stasis
US Initiatives Initiative 1 - Augmenting Operational Exercises with Formal Experimentation Initiative 2 - Integration System Program Offices and New Certifications Initiative 3 - Enhanced T&E Regime Earlier, Evidence-based Rigour and Innovation: Test Smart not Test Often! Initiative 4 - T&E Network Infrastructure Initiative 5 - Cybersecurity Protection Plans and T&E Initiative 6 - Permeating these U.S. Initiatives into Industry
Australian Recommendations Organisational. Experimentation exercises. Acquisition Policy. T&E Policy.
Introduction Challenge 1 - Growing System Synthesis Challenge 2 Higher-Order Human Functions Challenge 3 Cyber Threat Complexity response to Information Dominance Challenge 4 - Requirements Stasis
Change Failures due Disruption "I think there is a world market for maybe five computers." Thomas Watson, President IBM (1943) Kodak-Eastman built one of the first digital cameras in 1975. The Internet will soon go spectacularly supernova and in 1996 catastrophically collapse." Robert Metcalfe, founder of 3Com (1995) We don't feel it's having any impact on our results or that it has hit our radar as of yet," Richard Jones, VP of US Hospitality Ventures Management Group (2014).
Successes due to Disruption Gulf War I need for bunker buster inspired innovative use for old hard-steel naval artillery gun barrels in storage with PAVE WAY IIIs on F-111Fs. RAAF AP-3C Orion employed in coalition overland ISR missions in Afghanistan using sensor systems originally designed for maritime patrol needs (which is not an approved USN role!). RAAF modified US C2 software (TBMCS) designed for a permanent fixed HQ, driven by need for field-deployed air operations HQ.
5 th Gen Planned Air Force 1 st Gen 2 nd Gen 3 rd Gen Jet Fighter Generations Bird, SQNLDR S., 2014, 6th generation counter air missile architecture, Cranfield University, UK, July 2014 used a modified version of Tirpak J.A. See http://www.airforcemag.com/magazinearchive/pages/2009/october 2009/1009fighter.aspx First Jets Subsonic Guns Bombs Rockets Swept Wings Radar Ranging IR Missiles Supersonic Pulse Radar BVR Engagement 1940 s 1950 s 60-70 s Pulse-Doppler High E-M/ High-G Look-Down Shoot-Down 4 th Gen Extreme AoA Sensor Fusion Reduced Signature AESA Radar Datalinks 4+ Gen 5 th Gen All Aspect Stealth Internal Weapons Extreme E-M & AoA Full Sensor Fusion Adv. Sensors Supercruise Adv Datalinks 6 th Gen Ultra- Supercruise? UCAV Wingmen? Broadband EA? STOL? Cyber? Directed Energy? Broadband Stealth? Extreme Range? Large Weapon Bays Fully- Networked? 70-90 s 90-00 s 00-10 s 20-30 s
Arc of possibilities
Disruptors/Drivers? Beyond Planned Air Force: air power drivers and disruptors in 2030+ Hypersonics Uninhabited Systems Stealth/Low Observables Autonomous Systems Artificial Intelligence Modularised Weapons and Payloads Directed Energy Weapons Multi-domain Surveillance Electronic Magnetic Spectrum Operations Virtual Presence Communication Security Quantum Technology Space Systems Improved Global Navigation Systems Energy Security Climate Change Demographic Shifts
Hypersonic air power
AI and existential threats Machine-learning is the ability of a machine to learn new abilities without having been explicitly programmed by it s designer(s). Artificial Narrow Intelligence (ANI): machine intelligence that equals or exceeds human intelligence for specific tasks: IBM s Deep Blue (Chess), Google s AlphaGo (go), High-Frequency Trading Algorithms, or any automatic systems performing beyond human reach (Google Translate; spam filters; the guidance systems of point-defense anti-missiles/cannons etc.). Artificial General Intelligence (AGI or strong AI ): machine intelligence meeting the full range of human performance across any task; and Artificial Superintelligence (ASI): machine intelligence that exceeds human intelligence across any task.
You and being in-the-loop Human-In-The-Loop (HITL) The system design provides the operator with the ability to activate and intervene, to update or terminate, the mission. Human-On-The-Loop (HOTL) A system that, once activated, is intended to autonomously engage the target selected by the human operator. The system design also provides the operator with the ability to intervene to change or terminate the mission. Human-Out-Of-The-Loop (HOOTL) A System that, once activated, can autonomously select and engage targets without further human intervention. Level of Autonomy Mission Decision Transitions Find Fix/Track Target Engage Abort Manual Human Human Human Human Human Human-in-the-loop Human or System Human Human System Human or System Human-on-the-loop Human selected?? System System System Human or System Human-out-of-the-loop System System System System Human
US Initiatives details Initiative 1 - Augmenting Operational Exercises with Formal Experimentation (Dr Mal) Initiative 2 - Integration System Program Offices and New Certifications (Luke) Initiative 3 - Enhanced T&E Regime Earlier, Evidence-based Rigour and Innovation: Test Smart not Test Often! (Dr Mal) Initiative 4 - T&E Network Infrastructure (Dr Keith) Initiative 5 - Cybersecurity Protection Plans and T&E (Dr Keith) Initiative 6 - Permeating these U.S. Initiatives into Industry (Luke)
Australian Recommendations Organisational. Experimentation exercises. Acquisition Policy. T&E Policy.