Modularisation of Open-Hardware to Tackle the Digital Winter
Room B | Wed 22 Jan 11:40 a.m.–12:25 p.m.
Presented by
-
Associate Professer Paul Gardner-Stephen is an open-source developer, humanitarian technologist, engineer and academic who has been working on solutions to freedom of civil communications since 2010, including founding the Serval Project, a system for resilient off-grid communications among mobile phones, with numerous publications in the field. Having spent considerable times in Pacific Islands, as well as having lived and worked in remote areas, Paul has a strong understanding of the challenges that arise when communications infrastructure is damaged, disabled, or simply missing. Paul is also a founder of the MEGA65 open-source retro-computing project, which is now a successful open-source product, with almost 2,000 MEGA65s now creating joy for 8-bit enthusiasts, with its FPGA recreation of the never-released Commodore 65 (not the Commodore 64) (see https://mega65.org and https://github.com/mega65), and continues to contribute to the hardware, software and FPGA work on this community-led project, including a hand-held variant.
Abstract
The Digital Winter is what happens when unrest, supply chain challenges or a general societal breakdown make it difficult or impossible to create new devices and things. Or to maintain the ones that we still have. The extreme result, would be a kind of Mortal Engines style scavenging of "Ancient Tech", to meet the needs of the ongoing society, but we'd much rather a better way.
Together with the NLnet Foundation, we are exploring exactly this concept, to make a mobile-phone like device, whose goal is to enable and sustain communications and basic compute functionality, even if the rest of the world goes to pot.
This creates challenges, and requires trade-offs: The device has to be secure, to survive in the modern cyber landscape, that is populated with state-level actors and other malicious parties. It also has to be simple enough, that an individual or small group can maintain it, reason about it and improve it. These two actually go together: Simplicity is required to limit the attack surface size, to make it defendable by small groups.
Modularisation and sand-boxing of key components and modules looks like a feasible approach to this, and also to solve the hardware supply chain problem, by making it much easier to replace hard-to-source components with others, and also allows distributed manufacture and assembly, at low risk. For example, tricky to hand-solder components can exist more or less alone on a small PCB: Failures in assembling those don't require you to debug a whole board or throw it away. Similarly replacing those individual subsystems to respond to supply challenges, or to swap out untrustable parts with safer ones becomes much easier.
I'll explain how I am combining these methods, together with applying my past work in post-disaster communications to create a device that should be cost-effective to design, build and operate, and that is independent of energy and communications networks, and the trade-offs required to achieve this, including Intentional Obsolescence and judicious use of Somebody Elses Problem fields.
The Digital Winter is what happens when unrest, supply chain challenges or a general societal breakdown make it difficult or impossible to create new devices and things. Or to maintain the ones that we still have. The extreme result, would be a kind of Mortal Engines style scavenging of "Ancient Tech", to meet the needs of the ongoing society, but we'd much rather a better way. Together with the NLnet Foundation, we are exploring exactly this concept, to make a mobile-phone like device, whose goal is to enable and sustain communications and basic compute functionality, even if the rest of the world goes to pot. This creates challenges, and requires trade-offs: The device has to be secure, to survive in the modern cyber landscape, that is populated with state-level actors and other malicious parties. It also has to be simple enough, that an individual or small group can maintain it, reason about it and improve it. These two actually go together: Simplicity is required to limit the attack surface size, to make it defendable by small groups. Modularisation and sand-boxing of key components and modules looks like a feasible approach to this, and also to solve the hardware supply chain problem, by making it much easier to replace hard-to-source components with others, and also allows distributed manufacture and assembly, at low risk. For example, tricky to hand-solder components can exist more or less alone on a small PCB: Failures in assembling those don't require you to debug a whole board or throw it away. Similarly replacing those individual subsystems to respond to supply challenges, or to swap out untrustable parts with safer ones becomes much easier. I'll explain how I am combining these methods, together with applying my past work in post-disaster communications to create a device that should be cost-effective to design, build and operate, and that is independent of energy and communications networks, and the trade-offs required to achieve this, including Intentional Obsolescence and judicious use of Somebody Elses Problem fields.