! function(o, t, e, a) { o._aoForms = o._aoForms || [], o._aoForms.push(a); var n = function() { var o = t.createElement(e); o.src = ("https:" == t.location.protocol ? "https://" : "http://") + "support.flex.com/acton/content/form_embed.js", o.async = !0; for (var a = t.getElementsByTagName(e)[0], n = a.parentNode, c = document.getElementsByTagName("script"), r = !1, s = 0; s < c.length; s++) { if (c[s].getAttribute("src") == o.getAttribute("src")) r = !0; } r ? typeof(_aoFormLoader) != "undefined" ? _aoFormLoader.load({ id: "482592df-d182-4ff0-be56-eebb313261fa:d-0001", accountId: "39314", domain: "support.flex.com", isTemp: false, noStyle: false, prefill: false }) : "" : n.insertBefore(o, a) }; window.attachEvent ? window.attachEvent("onload", n) : window.addEventListener("load", n, !1), n() }(window, document, "script", { id: "482592df-d182-4ff0-be56-eebb313261fa", accountId: "39314", domain: "support.flex.com", isTemp: false, noStyle: false, prefill: false });
All or nothing: How autonomous vehicle success is driven by cloud and edge computing

Cloud-based data storage changed the way we interact with data and applications. It solves many complex problems, but it’s not a singular solution. Autonomous vehicles (AVs) and other applications that require a huge amount of data need super-fast analysis and delivery. They can’t rely on cloud-based computing alone.

That’s why we include edge processing. It reduces latency by positioning computing closer to the application. This capability is essential in designing for autonomy. Supporting edge computing is critical to supporting AVs.

Here's how the cloud and edge come together to power autonomous vehicles.

The edge processes urgent information in near real-time speed

As an AV travels down the road and through traffic it encounters rapidly changing road, weather and light conditions, other vehicles, pedestrians and animals, traffic signs and more. It gathers data about each of these factors using photography and video, radar/lidar, accelerometers and telemetry and gyroscope sensors.

The data is processed in the vehicle (autonomous compute), an edge computer (micro data center or nearby device) or the cloud (today’s typical data centers). Data that is critical to navigation is processed directly in the vehicle and then is sent to other processing locations based on level of urgency. This initial analysis also incorporates machine learning so the vehicle can continually learn and adapt to its surroundings.

At the same time, data gathered from the vehicle is shared with the surrounding infrastructure to support both real-time analysis and long-term planning and improvement.

The processing delivers feedback to the vehicle’s navigation controls so it can make informed decisions. That analysis needs to happen as quickly as possible so the vehicle can keep up with constantly changing driving conditions. Edge computing creates the connected gateways needed to effectively complete this cycle.

Cloud providers are investing and building out edge infrastructure to make this picture for time-sensitive information processing a reality.

The cloud’s role is to process and store complex historical data

A traveling AV also creates and stores a record of its activity. Since historical data is not time- or safety-sensitive, it is uploaded from the vehicle to the cloud at convenient and cost-effective times. A typical opportunity is when the vehicle is parked overnight and connected to Wi-Fi. Aside from historical data, the cloud also has capacity to handle larger, more complex workloads. For example, neither the edge nor the vehicle can encompass the complex obstacles found in managing a fleet of cars. That’s where the power of cloud computing comes in handy.

Huge amounts of historical data are stored by AV manufacturers. When first captured and compressed for storage, data about a single trip may appear to have very little value. But as manufacturers’ autonomous driving applications advance, the data creates information that can be used to enhance both AV design and infrastructure enhancements.

While the cloud handles complex processing and storage, edge computing continues to react instantly to surroundings and actions. The cloud collaborates with the edge to smartly and efficiently manage data.

Cloud and edge computing come together to support AVs

AVs have the potential to change the way we travel. To navigate safely and become socially acceptable, AVs must be backed by an infrastructure with impeccable data gathering, communication, analysis and management. Cloud computing, edge computing and a well-oiled data center are crucial pieces of that infrastructure.

Urgent processing takes place within the vehicle’s edge computing center. The vehicle concurrently sends data to be used by other vehicles and other pieces of the supporting infrastructure. The cloud gathers the whole mass of data. This is how edge and the cloud work together to manage the complexity of AV navigation.

There is a high level of market potential in AV support, as well as 5G and other autonomous applications. Data networks can’t support the new level of autonomous speed and complexity if they only utilize the cloud. The writing is on the wall: rapid expansion of edge and cloud data centers is around the corner. Is your business ready?

Explore how our edge computing and autonomous vehicle capabilities can help you.


Related articles

Autonomous vehicles have a lot to learn
Computing power for AV success. For an autonomous vehicle (AV) to learn the complex task of driving, you need cloud and connectivity.
Flex Talks schedule – medical devices
At a glance
Flex Talks webinar schedule – learn about Ideas that are transforming medical device design and manufacturing.
Infrastructure for essential services
Flex is partnering with our cloud and communications customers to manufacture essential medical equipment.

Subscribe to our blog

You can learn more about our use of cookies and your choices here. By clicking on any link on this page or clicking "OK, I agree", you are giving your consent for us to set cookies.