Innovation Monitor: A deep dive into Edge Computing
Welcome to this week’s Innovation Monitor. This week, we’re deep diving into the world of edge computing. Let’s begin with a definition:
Edge Computing: “a part of a distributed computing topology in which information processing is located close to the edge — where things and people produce or consume that information.” — Gartner
The term is most closely associated with IoT devices — such as internet-enabled cameras, smart speakers, and yes, devices like Alexa and Google Homes. However, this reflects confusion, not reality, around the role of edge computing in a tech stack. Common questions include: Aren’t these devices already “computing”, though? Is edge computing just adding more compute to existing IoT devices so they process stuff faster? And why don’t you just send all that sensitive data to the cloud, since that’s what we’ve been doing for the past decade?
Well, we’re already privy to nefarious actors using Google Home and Alexa to eavesdrop and phish passwords; Amazon workers themselves listen in on conversations, in addition to the many IoT botnets circling around.
Conversely, an edge device that can process data locally is technically less prone to attack or vendor intrusion. In addition to the privacy benefits, devices that can compute complex calculations on the edge can also save companies on bandwidth, cloud processing, and storage costs.
So there’s a demand for fast and intelligent local systems. As Ars Technica noted in Here’s why Apple believes it’s an AI leader — and why it says critics have it all wrong: “…increasingly, Apple performs machine learning tasks locally on the device, on hardware like the Apple Neural Engine (ANE) or on the company’s custom-designed GPUs (graphics processing units). Giannandrea and Borchers argued that this approach is what makes Apple’s strategy distinct amongst competitors.”
Infographic credit: Network World
It’s also helpful to think of edge computing as part of an ecosystem of tech. That ecosystem can consist of edge devices (the IoT device itself, edge servers, even a powerful smartphone), device/IoT management applications, cloud services like AWS, AI, and 5G connectivity.
The main takeaway, for now, is that edge computing can help companies beef up data privacy and security, reduce bandwidth and compute costs (and increase resilience in case of an internet outage), and enable real-time applications.
As always, we wish you and your community safety, calm and solidarity as we support each other through this unprecedented time. Thank you for reading!
Erica Matsumoto Related Innovation Monitor Editions
As we’ve been writing this newsletter for nearly two years (wow) and have an entire archive, we’ll include a brief section linking to past editions that provide more context to this week’s topic:
August 22nd, 2019: The Future is 5G
October 21st, 2020: Beyond Silicon (examining the Gartner Hype Cycle trends) Edge Computing Ecosystem Use Cases Now for the fun part — what can this ecosystem of tech enable? Here’s a quick rundown, and feel free to click on any of these for a more in-depth exploration. We’ll explore 5G’s role (and the mysterious fog computing) in the next section.
Newer cars have tens of millions — if not over 100M — lines of code. Autonomous vehicles (and really most vehicles going forward) will require real-time processing for split-second decisions. While heavy, vital processing happens on edge, a connected car might sync with your other devices for a more personalized experience (“changing the mirrors, giving you an update from your calendar of your schedule, and lining up your iTunes playlist”). Cars will also be able to communicate with one another. As Wired writes, “on the road, cars will talk to each other, automatically transmitting data such as speed, position, and direction, and send alerts to each other if a crash seems imminent.”
Yes, the space is cratered with the downfall of Sidewalk Lab’s Quayside project, among other failures. But other city-from-scratch projects push forward, such as Egypt’s planned smart city 28 miles east of Cairo. Data collected from the city’s smart buildings is expected to cut energy consumption and meet sustainability goals. Neom, Saudi Arabai’s supersized smart city, is expected to have “flying cars and robot dinosaurs.” But smart city projects don’t have to be so far-reaching. Buildings can be built (or retrofitted) with edge devices that adjust power, lighting, water, and ventilation systems to save costs and reduce environmental impact (check out Bloomberg’s European headquarters for a cool example).
In 1980, the cost of solar hovered around $30 per watt — today, this price has gone down to about 30 cents, making it affordable for the average home. While a smattering of rooftop solar panels in Brooklyn won’t shift our reliance on traditional grids, the installation of batteries might help. Tesla’s Powerwall battery, for example, stores energy and can be used as a backup power source during an outage. Hook a neighborhood of panels and batteries up, and you have a local virtual power plant that can feed power back to the grid (compensating homeowners) and prevent blackouts. Sensors and IoT devices would be a key component (in factories, plants, offices, or homes), connecting to an edge platform to monitor and analyze energy consumption in real-time.
Agribots & Farm Automation
Autonomous tractors have seen significant progress over the past several years. Earlier this year, Kubota launched a completely autonomous tractor that is able to make its way around both regular fields and rice paddies; John Deere showed off a concept electric tractor at last year’s Agritechnica trade fair; and a group of researchers at Harper Adams University is planning to start harvesting crops autonomously this year. Eastern Peak writes that autonomous tractors and other robotic farm devices (for, say, watering, weeding, or soil inspection) can communicate with nearby sensors to gather data about the surrounding environment. Edge computing is especially important in rural areas with limited internet connectivity.
And just for kicks, here’s a DIY facemask detector for under $200 using an Nvidia Jetson — a small computer that can be built into drones, robots, medical equipment, or any other device. (Also check out McKinsey’s 107 user cases of edge computing.) 5G & Fog Computing Edge computing offloads compute to local devices, saving on bandwidth and reducing latency. But for some use cases, like cloud gaming, virtual reality, and augmented reality, decreasing hardware load is preferable — at least, until the hardware is small enough.
For example, fitting a powerful computer capable of projecting a digital overlay over the world in a pair of lightweight glasses is not feasible (AR is hard). But 5G has years left to mature. A compromise can be reached until 5G is more uniqutous, with edge computing and 4G creating a similar experience.
Image credit: The Fast Mode
As The Fast Mode puts it: “Today, there are only nascent markets for the types of applications 5G enables: augmented reality, mass IoT, robotics, AUVs/drones, etc. Edge computing can provide developers an environment to create the 5G applications that do not exist today even without ‘full 5G’ being available yet.”
But edge computing isn’t just a convenient bridge to 5G’s future potential: “Edge can enable operators to change their backhaul business models. For data-heavy applications, such as those requiring high-definition video or extensive data analysis, even with 5G, sending data constantly back to the cloud will be expensive and deteriorate the customer experience. Instead, data could be filtered out, with the full stream travelling only as far as a local edge site, before being analysed, rationalised, and only what is necessary streamed and stored in the centralised cloud.”
Image credit: The Fast Mode
Telco’s see 5G and edge devices as an opportunity for hosting local applications. Netflix, for example, has been partnering with local ISPs via its Netflix Open Connect program for years to help offload heavy video streaming traffic (source: EE Times). 5G providers could offer similar services for streaming content, gaming, XR, and other applications.
Image credit: EE Times
Let’s wrap things up with fog computing. As spooky as it sounds, the Cisco-coined term is pretty the same as edge computing, though “there are many who delineate the fog either above or below the edge computing space or even as a subset of edge computing.” So with unintentionally amusing phrases like, “the global fog market has the potential to reach $768 million USD worldwide by the year 2025,” just replace “fog” with “edge”. This Week in Business History This Week in Business History
November 12th, 1983: Cabbage Patch dolls are launched in Hollywood, California
Created by Xavier Roberts and licensed for mass marketing to Coleco Industries, each doll was physically unique and came with its own birth certificate and adoption papers. The toys created such demand that they were one of the first real “Black Friday” shopping crazes in the year of their launch.