Plenty of people have talked about bringing common order and interoperability to the fragmented and insecure ecosystem that is the internet’s edge. They’ve dreamed of introducing a common architecture to a heterogeneous, distributed technical landscape of automotive computers and smart fridges; telecoms base stations and wind farm diagnostics tools: all chattering away to their owners and vendors over a disconcerting array of industry-specific legacy connectors and recently introduced wireless protocols
The Linux Foundation (LF) this week took a large step towards making this aspiration a reality with the launch of LF Edge; an ambitious umbrella project dedicated to creating an interoperable open framework for edge computing. LF Edge – to be compatible with any hardware, silicon, cloud, or Operating System (OS) – has attracted more big-name sponsors than you could shake a stick at; from Arm to AT&T, via Dell, Ericsson, HPE, Huawei, IBM, Intel, Juniper Networks; Qualcomm, Red Hat, Samsung, and many more.
To understand the project, think Android, but for the Internet of Things (IoT), and you would not be far off: i.e. an open architecture that decouples applications and the infrastructure they’ll run on in order to allow development of cloud-native apps for a sweeping range of both consumer and Industrial Internet of Things (IIoT) devices.
Picture the ability to push new tools and functions from the cloud to IIoT/edge hardware: e.g. advanced analytics updates to the box harvesting your wind farm data.
LF Edge: 60 Founding Members, 5 Projects, 1 Goal
LF Edge, unveiled on Thursday (and boasting an impressive 60-strong list of global founding members) comprises five discrete projects, aimed at bringing a silo-busting framework to the consumer, industrial and telecommunications edge.
Linux Foundation General Manager Arpit Joshipura said: “The market opportunity for LF Edge spans industrial, enterprise and consumer use cases in complex environments that cut across multiple edges and domains.”
He added: We’re thrilled with the level of support backing us at launch, with more than 60 global organizations as founding members and new project contributions. This massive endorsement, combined with existing code and project contributions like Akraino from AT&T and EdgeX Foundry from Dell EMC, means LF Edge is well-positioned to transform edge and IoT application development.”
The Five LF Edge Projects
Project 1: Project EVE (Edge Virtualization Engine), contributed by startup ZEDEDA, is essentially an Operating System for the IIoT edge that aims to decouple application development from infrastructure solutions, underpinned by hardware-assisted virtualisation.
Project 2: Akraino Edge Stack is an OSS stack for cloud services optimised at the edge running in VMs or containers. With many similarities to EVE, its focus however is deployment at the edge of telecommunications network, e.g. base stations. Code is coming in early 2019: its seed code was contributed by AT&T.
Project 3: EdgeX Foundry is a common edge microservices layer. Its aim is to boost interoperability by letting adopters deploy a mix of plug–and–play microservices on compute nodes at the edge; e.g. to support analytics, data orchestration, database, security, system management and services. It ideally needs EVE or Akraino to run, but users can also get started with Linux and Docker runtime.
Project 4: Home Edge Project, with seed code contributed by Samsung Electronics, is a consumer edge computing services platform, for devices from smartphones to fridges. It will come with a set of APIs that can also run with libraries and runtimes.
Project 5: An Open Glossary of Edge Computing provides a concise collection of terms related to the field of edge computing.
Computer Business Review spoke to the co-founder of ZEDEDA, Roman Shaposhnik, who has been working on LF Edge’s IIoT project EVE since 2017.
Shaposhnik, who also sits on the board of directors of the Apache Software Foundation explained: “Android is a good analogy: in the past if you liked the functions or applications that came with your phone, when you changed phone manufacturer, those were lost. It was all vertically integrated. Now, of course, your apps are decoupled from the device. So the aim with EVE is to let developers extend cloud apps to edge devices without the need for specialised engineering tied to specific hardware platforms.”
“This isn’t easy. For the enterprise edge, locations are highly distributed, the hardware is on-premises and very diverse”.
As Shaposhnik puts it: “Wind farm analytics are a good example. At the moment if you are managing a fleet of turbines then you need and want to do predictive maintenance, as it is expensive to dispatch an engineer when things break. But vibration analytics generate something like 1GB of data per second. It’s too much to send to the cloud, so they have to analyse it locally; with the edge there’s an additional forcing function because the levels of data generation force a hybrid solution.”
“But you end up with all these vertically integrated solutions. Customers ask why you can’t do shootouts of applications like you do with a smartphone and you just can’t: everything is siloed: you have all these boxes on your wind farm, all ultimately vertically integrated to their own hybrid cloud.”
“So basically Project EVE is about breaking vertical siloes into reusable components, creating an architecture in which apps become true software deliverables.”
With the number of edge devices expected by Gartner to exceed 20 billion by 2020, in order for the broader IoT to succeed, the currently fragmented edge market needs to be able to work together to identify and protect against problematic security vulnerabilities.
The rise of containers and hardware-isolated micro-virtualisation technologies has made this more possible than ever. Shaposhnik told us: “We’re building a hardware root of trust – just like Android – and using trusted execution environments on Arm etc.
He adds: “In a data centre you don’t need to worry quite so much about physical attacks: you’ve got a guard at the front with a gun. At the edge physical security is a paramount concern. Can we make sure devices we deploy are as difficult to crack, even if you have physical possession of them, as an iPhone? That’s what we’re doing. We’re extremely excited to be part of LF Edge.”