How edge can help improve sustainability -- and how it can't

Exploring edge's sustainability through real-world research

Edge computing refers to a model of application deployment that puts either part or all of the application hosting close to the systems or operations it's meant to support. It typically uses a small server with limited system software and performance that is connected to other software that runs on the cloud. It requires high reliability and low latency to deliver on its promise of real-time insight.

Most articles that examine "benefits and downsides of edge computing" do so with only generalities. Instead, this article draws from firsthand internal research with real-world enterprise users.

Of 420 enterprise leaders who shared their compute policies with Andover Intel, 307 were using some form of edge computing. Of those using edge, 210 were not technology enterprises, and these will be the focus for the rest of the article. In about half of these cases, leaders characterized their applications as IoT rather than edge computing. However, all the IoT initiatives used edge computing to handle IoT events and control processes. What these real-world users shared is the basis for the views and statistics in this article, and they can shed light on the ways in which edge computing can help support -- or hurt -- sustainability.

Of the 210 enterprise edge users, 202 said that sustainability was a goal in their edge deployments, but only eight said it was the primary goal. The primary goal was reserved for various aspects of cost control. The mix of project justifications demonstrates that it is possible to make a traditional cost/benefit business case for sustainable technology decisions.

Still, enterprise edge computing users were unanimous in saying that edge computing technology was one of, if not the most important element, in implementing a sustainability project.

As an illustration, of the 38 who indicated that they relaxed the normal financial business case constraints to enhance project sustainability, 30 said that edge computing was critical in giving them a viable project that was also sustainable.

Edge computing's potential sustainability benefits

Reducing water and energy consumption, and carbon emissions is a major challenge for most companies. To meet sustainability goals, more companies are turning to low-latency applications to control processes. This capability has been one of edge computing's main selling points.

Energy reduction

One way edge computing has the potential to help support environmental sustainability initiatives is through its ability to facilitate data transfer with less energy usage both at the macro level and for specific organizational efforts.

Power grids are one demonstration of this functionality.

As a result of public awareness of sustainability and government support for initiatives, the power grid is becoming more decentralized to accommodate a multiplicity of energy resources. The five electric utilities that responded to Andover on their edge computing use reported relying on edge computing to adapt to such changes.

Organizations of all types are also turning to edge computing to help manage -- and lower -- energy consumption. In fact, 187 of the 210 edge users reported to Andover that reducing electricity, gas or oil costs was a contributing justification for the edge project.

More sustainable logistics

Edge computing shows promise in supporting more sustainable logistics.

Three companies that introduced electric vehicles to their fleet say that edge computing is essential in ensuring optimum availability via vehicle dispatch and charging management. Policies determining the vehicle assignment to tasks and routes to maximize range were cited by all three.

The use case shows up in traditional gas and diesel logistics, as well.

Nine of the 12 road transportation companies report using edge computing for assigning and routing gas or diesel vehicles to save fuel.

Waste reduction

Edge computing can help reduce process waste from almost anything.

Of the companies that were from farming, manufacturing, mining or petroleum industries, a majority said they use edge computing to reduce resource waste.

For example, all four of the agricultural giants that responded about their edge computing strategies reported using edge to help manage irrigation control through the use of soil and crop sensors to monitor conditions, crop type and phase of growth and weather forecasts.

There is also growing support for the use of edge computing in waste management in large company facilities and university campuses.

In this use case, organizations are turning to IoT and edge computing to scan waste and look for items that require special handling, such as e-waste. Of 22 such facilities leaders who offered Andover Intel comment on edge computing, 12 had implemented or were implementing this edge application.

In manufacturing, waste reduction can help save on costly raw material and component parts, and control the amount of material, including e-waste for technology products, that might otherwise end up in landfills or require costly reprocessing for disposal.

While waste control isn't typically reported as a project justification, 11 of 28 manufacturing leaders cited it as a sustainability benefit, a rate only slightly below that of energy savings and emission reduction. Six of the 13 companies that started their processes with raw materials reported using edge computing in waste control. Five of the 15 companies whose manufacturing involved only assembly reported using edge computing for waste control.

Smart buildings and cities

Smart buildings rely on edge computing and IoT applications to optimize fuel usage and minimize emissions. In these cases, edge helps control heating and cooling in off-hours or when no personnel are present or activities are happening.

Of the 210 enterprises that reported edge computing use, 94 had at least one smart building where edge computing controlled energy use.

Smart city implementations are also using edge computing to control traffic signals and reduce congestion, which reduces emission by cutting down on idling in slow or stopped traffic. This strategy seems to be of the greatest interest in heavily congested metro areas such as in China and Europe, and involves the use of edge computing to model traffic flow based on sensor detection.

Officials from two cities reported this edge computing use, though in both cases, the project is still in the trial phase.

Edge computing's sustainability challenges

For all of edge computing's promise in supporting sustainability, the technology also has a number of negative impacts. Here are a few of those.

Discounted negative sustainability impacts

As with the use of AI for sustainability, the hype surrounding edge computing has given it a reputation as a panacea. In turn, too many IT leaders fail to account for edge computing's negative environmental impacts.

Edge computing -- as with any organizational or digital technology -- is not an abstract concept. It requires hardware -- for example, of sensors -- initial manufacturing and so on. Edge computing setups -- such as IoT sensors -- require raw materials, manufacturing and transportation. And edge computing setups also use energy and create end-of-life e-waste.

While edge computing has the potential to support sustainability, that potential must be examined to determine whether it's net negative or positive.

Tendency to become tech debt

Edge computing can only support efficiency -- and in turn, sustainability -- if organizations use it.

However, more than 90% of the 210 enterprises who offered Andover Intel comments on edge computing said that their edge computing resources are the most underutilized, and therefore the most wasteful, of all their facilities. None said they were moving work from the cloud to the edge.

The need for dedicated space

Edge computing setups need to be close to the processes they're controlling. That means they usually share space with those processes, not offices or data center space.

In other words, edge computing almost never capitalizes on existing data centers and their power and cooling capabilities. Instead, IT teams tend to create new facilities to house edge computing operations, because of the low-latency goals.

Of the 210 enterprises that offered edge computing comments based on their own operations, 54 said that to use edge computing required new dedicated space for installation or operation. In 13 cases, the cost of this facility in energy terms, for a proposed project, was high enough to compromise the project business case to the point where special approval was needed, and in four of the applications, the approval was not forthcoming.

High energy requirements

Especially since edge computing often facilitates other technologies -- such as AI or digital twins -- edge computing can be a gateway to higher energy requirements.

To this point, some enterprise leaders are concerned that the evolution of edge computing technology will increase the compute power needs at the edge and raise the sustainability burden. For example, of the 210 companies reporting edge-computing use, 11 indicated that they had trialed the use of what they called digital twin modeling or AI in an edge application and found that it doubled energy requirements.

E-waste

Edge computing's use of IoT sensors is a particular source of e-waste, sometimes a significant one.

Leaders from 54 enterprises reported that they were surprised, even alarmed, at the rate at which sensor or controller IoT elements in their edge applications had to be replaced. One company installed over 200 sensors, only to find that half had failed within a year.

The key to increasing IoT investment longevity is to carefully review whether the sensors in question can withstand the environmental conditions where they'll be placed and be willing to pay a premium for devices whose protection from environmental hazards offers a comfortable safety margin. Otherwise, sensor failure waste, and in turn, greater amounts of e-waste, will be the reality.

Public edge skepticism

Many of the negative impacts edge computing could have on sustainability could be reduced by the use of public edge computing services -- in theory. For example, inefficient use of power and cooling for edge computing could be reduced if public edge services, similar to public cloud services, were available.

In practice, leaders from only four of the 210 companies using edge computing could envision switching to a public edge or cloud service. Most believe that the latency requirements and availability requirements of their applications could only be met through hosting edge applications themselves, local to the processes being controlled.

While 5G -- and now 6G -- technology has promised to create low-latency connectivity, only 28 enterprises currently use either, and only 5G is currently available. And these enterprises used 5G only for a small portion of their edge needs. Lack of credible public edge services to back up local edge hosting is seen by over 130 of the edge users as a barrier to edge computing adoption.

Analysis of sustainability and edge computing ties

Every technology decision made by enterprises is made based on the potential for a positive ROI, which balances return against cost. Edge computing relies on the same principle, as do sustainability initiatives.

The good news is that almost all sustainability goals involve a consumption reduction -- and therefore cost -- of some resource a company pays for. In other words, sustainability initiatives tend to have an inherent positive ROI promise. Power consumption demands power generation, so saving the former reduces demand on the latter, which is a sustainability benefit. Reducing the use of field personnel for maintenance and support reduces the cost of drivers, vehicles, insurance and fuel, which in turn, lowers fuel consumption and emissions.

IT leaders can do some power usage optimization through planning. But to have the greatest impact they have to control activities in real time, which demands a short control loop, meaning a small delay between the sensing of a condition and the response. Edge computing achieves that due to its nearness to the processes that generate environmental impact and which demand control responses.

On the other hand, in terms of computer use, edge computing is not efficient because it has to be highly reliable and available at a point where it's unlikely it can share the edge resource with other processes and also meet latency demands.

CIOs and others seeking environmental sustainability improvements can look to edge computing, albeit with the downsides weighed against the benefits, as one way to help support change.

Tom Nolle is founder and principal analyst at Andover Intel, a consulting and analysis firm that looks at evolving technologies and applications first from the perspective of the buyer and the buyer's needs. By background, Nolle is a programmer, software architect, and manager of software and network products, and he has provided consulting services and technology analysis for decades.