Energy Autonomy Needed for Sustainable IoT Infrastructure

Article By : Shohei Kawanaka

Why deploying a large-scale IoT network is taking longer than many would have hoped.

There has been no shortage of spokespeople evangelizing about the Internet of Things (IoT) in recent years. They have talked about the huge potential of this technology and the multitude of benefits it can offer to our society — more efficient industrial processes, addressing traffic congestion, boosting crop yields through smart agriculture, improving patients’ quality of life via digital health, etc. Though all this sounds positive, it cannot be denied that large-scale IoT network rollout is taking longer than many would have hoped.

There are still several obstacles present here, some relating to the technologies employed and others that are more logistically based. Consequently, the implementations predominantly witnessed so far have been on a relatively small scale. The following article looks at what can be done to overcome these obstacles so that IoT infrastructure, which is streamlined, cost-effective, and sustainable, can be created.

What is currently holding back IoT adoption?

There is a number of reasons for the slower-than-expected progression of IoT implementation. Among the most prominent of these are the following:

  • The need to lower associated hardware costs. Most IoT nodes require inclusion of many different components, and this has cost implications. By curbing the upfront financial outlay in infrastructure, willingness to adopt IoT technology is certain to increase. The number of enterprises/municipalities/organizations able to roll out IoT networks would then broaden, leading to a more expansive array of use cases being addressed.
  • IoT node size. The large quantity of components needed in IoT hardware raises another issue. As well as the sizable bill-of-materials (BoM) costs, it also means that nodes can be quite bulky. This precludes them from being used in a significant proportion of potential applications in which space constraints must be taken into consideration (such as ones that are body-worn or needing to fit into tight enclosures). Therefore, ways to contract node dimensions need to be found but without limiting these nodes’ functional capabilities or shortening their operational lifespan (due to them not having enough energy storage available).
  • Minimizing operational expenses. It must be recognized that implementing an IoT network is just the beginning. After this, continuous effort is required to maintain the operation of all the nodes comprised.
  • The levels of e-waste the IoT network will be responsible for. This is another important point that cannot be overlooked. As we will see, IoT networks can have a dramatic environmental impact unless appropriate measures to prevent this are taken.

Before embarking on IoT network implementations, it is important that organizations understand the total cost of ownership that a network will represent. The first and third points listed above should be combined to give an accurate estimate of what the overall costs will be over the network’s projected lifespan. Only then can it be assessed whether the investment will be justified.

Internet of Things - Sustainable IoT

Power concerns

How nodes are powered contributes greatly to an IoT network’s total cost of ownership, affecting both the BoM and the ensuing operational expenses. At the moment, the vast majority of deployed IoT nodes are dependent upon Li-ion battery cells as their power source.

Battery cells not only take up board real estate of IoT nodes, precluding them from the more space-limited applications already outlined, they also have other drawbacks. Battery charge will, of course, be used up over time.

The period that deployed hardware can be utilized for will be dictated by its power budget. If an IoT node needs to make many data transmissions per day, or has a large processing workflow to deal with, then this will be higher. Its battery will thus last for a shorter amount of time, and the battery will need replacing sooner.

If an organization’s network consists of tens of thousands of nodes, battery replacement will be continuous, calling for a dedicated maintenance team to be assigned to this task. This will severely inflate network upkeep expense. It will be exacerbated still further, if not made completely unworkable, should the IoT nodes be in remote locations (such as on petrochemical pipelines or in smart agriculture installations). Under such circumstances, the nodes will be too difficult to get to.

The environmental aspect

According to UN figures, approximately 50 million tons of e-waste get generated each year, with only a relatively small percentage of that (most estimates stating 15% to 18%) being recycled. Batteries represent a large proportion of the e-waste produced. Though there is ample recycling opportunity for other battery chemistries, Li-ion battery-recycling activity is still very limited. The number of Li-ion batteries currently being dumped is already proving problematic, but as IoT nodes become more prevalent, the number of batteries being disposed of annually will ramp up exponentially.

Most industry analysts forecast that overall IoT deployments will soon constitute a colossal number of nodes. Statista, for example, stated in its most recent report on the subject that there will be 29 billion IoT nodes in operation by 2030. Based on this (given an estimated battery lifespan of 18–24 months), 10–15 billion Li-ion cells’ worth of e-waste could be generated by the IoT alone each year.

Another contributor to e-waste that should be mentioned is the plastic remaining from the credit-card–sized carriers used for the subscriber identity modules (SIMs) of cellular IoT nodes. By replacing conventional SIM cards with directly integrated alternatives, this plastic buildup could be avoided.

Following an energy-harvesting strategy

From both a cost and environmental perspective, it is clear that powering the IoT by just using Li-ion cells will be impractical, and the situation will only get worse as the scale of IoT networks increases. This is why energy harvesting is proving of such interest in an IoT context. Rather than having a depleting electricity storage reserve, energy can instead be extracted from the ambient environment on an ongoing basis. In some cases, thermal gradients or vibrations will be the energy source; in others, it will be incident that is used.

Leveraging energy-harvesting technology, as opposed to battery-powered arrangements, will reduce IoT network costs considerably — eliminating the purchasing of new battery cells alongside the effort of replacing them. It will also have environmental plus points, with no e-waste being generated through battery-cell disposal.

Until now, uptake of energy harvesting in an IoT context has been to some extent hampered by constituent component costs. The power management ICs (PMICs) responsible for supervising the energy-harvesting process are at the heart of the problem. These devices make sure that optimal current is provided to the IoT hardware and that sufficient amounts are directed toward the rechargeable reserve (such as a supercapacitor) for later use. Unfortunately, such PMICs normally require support from a large number of passive components (normally at least a hundred). Likewise, the wireless modules available for use in this hardware are often too large and expensive to make their sourcing feasible.

Conclusion

There is no doubt that IoT infrastructure rollout is being slowed down because of fundamental implementation issues. The number of components needed is making IoT nodes too pricey for them to address many prospective scenarios. Also, the resulting node size is often too big for them to be effective. Through use of the platform described above, better-optimized IoT nodes can be produced. These nodes will not be dependent on batteries, with lower BoM costs, smaller form factors, and less environmental impact all being among the key advantages realized.

 

This article was originally published on EE Times Europe.

Subscribe to Newsletter

Leave a comment