The evolution of smart meters: Towards a supersmart grid

The Apple iPad could become the device of choice for consumers to monitor and control their energy consumption Source: Apple

The benefits of smart meters are well documented, but great strides are being made in smart meter technology. Willard Tu, embedded segment marketing manager at UK-based semiconductor intellectual property supplier ARM, examines the technological evolution of smart metering.

Willard Tu, ARM, USA

Smart metering, and the technology behind it, has become increasingly sophisticated since it was first brought to the attention of governments and end users during the last decade.

The benefits of smart meters are well documented, but the development of smart meter technology, required to keep pace with the increasing demand, is taking place quietly in the background. As demand for smart infrastructure increases and the move towards national Smart Grids gathers pace, the technology used for smart metering is evolving significantly to keep up.

Smart Meters Past and Present

The evolving nature of smart metering technology has brought about the increased need for 32-bit computing to facilitate faster, more sophisticated and more complex processing. Electronics in meters started life as a method of enabling Automated Meter Reading (AMR), which was a way for energy companies to reduce costs by reading meters more efficiently with a hand-held device or a vehicle driving down a neighborhood street.

When initially developed, such systems were designed with significantly less CPU power, typically 8-bit or 16-bit CPUs, and significantly less memory within the page memory boundary for those architectures, at less than 64 KB.

As meters were developed to meet the Advanced Meter Infrastructure (AMI), an increase in their processing capabilities was required. These meters were now capable of mesh networking, allowing the creation of a local area network (LAN)/wide area network (WAN).

Such mesh networks allowed the metering infrastructure to have a data collector hub that would then transmit the collected data via GSM, broadband or another means from an end user to the utility office. This customer data was then transacted automatically into customer billing. The AMI capable meters also typically included a Home Area Network (HAN), such as Zigbee. A Zigbee radio software stack, just as mesh networking does, needs increased CPU requirements of the MCU, and also larger memory.

The typical memory footprint to support RF LAN/WAN was around 256 KB. Comparatively a Zigbee was around 128 KB of embedded flash memory. In addition, energy companies wanted to incorporate field upgradeability.

Using the same mesh network to provide firmware updates to smart meters, this feature would require MCUs to have dual banks of flash memory. This dual memory bank architecture would allow the meter to run continuously, as one bank of memory is re-programmed with new software, enabling the meter to transition operation to the updated program. Suddenly, a smart meter’s MCU memory requirements went from less than 64 KB to greater than 640 KB and CPUs from 8-bit and 16-bit to 32-bit.

Motivation for Meters

The Smart Grid, as it is often referred to today, is merely an implementation to help energy companies manage their business needs. It is often portrayed as a benefit to the consumer. However, the many benefits that have been talked about are yet to be realized by the end user. Time-of-use payment plans and optional load shedding control by utility companies are frequently referenced as benefits, but have not reached the average energy consumer.

Smart meters can enable the customer to monitor the cost of energy with an integrated Home Area Network Source: ARM

Consumers identify with products that they actually interact with in their daily lives, such as appliances, TVs, computers and thermostats. A consumer will rarely look at or consider what a meter does for them. They simply act as a point of reference for most.

This is changing as the energy companies roll out the smart meters, and consumers are beginning to wonder what this new meter will do for them. These meters give the energy companies the ability to eliminate the service personnel that currently oversee activation, reading and disconnection of electricity usage. While this may be more convenient for the average consumer, it is more of a benefit to the energy companies in cutting their operational overheads.

So what do these smart meters do for the consumer? Some have the ability to enable the customer to monitor the cost of energy if there is a Home Area Network (HAN) integrated into the meter. The HAN might be wireless, such as Zigbee, or wired, such as Power Line Communications (PLC).

There are many other possible HAN communication protocols available. It will certainly help the customer to understand what the meter will do for them if the meter provides the means to communicate or interact with all of their consumer products.

Some examples are the new tablet devices, such as the Apple iPad, which is Wi-Fi enabled (although not using a Zigbee) and could become the device of choice that the consumer uses to monitor and control their energy consumption.

Many companies already have apps that would run on an iPad to enable home automation. One such company is Control4, which manufactures a ‘home control’ appliance that allows consumers to operate a variety of devices. Light switches, networked thermostats and touch-screen keypads, as well as televisions and home security systems, can all be operated from a single device.

So does the meter need to have Zigbee, PLC and Wi-Fi? This would allow the meter to potentially provide the function of a communication hub for the home: a “home gateway”. This communication hub could become integrated into future network routers, in set top box (STB) devices or as its own separate dedicated appliance.

The need to be flexible in communication with many different appliances, as well as the utility home office, will require the smart meter to evolve into a smarter meter and eventually to a ‘supersmart meter’. The software complexity will soar, in the same manner many other industries have experienced.

Processing and memory requirements will rise significantly beyond smart meters as implemented today, with approximately 640 KB of flash memory. This is already ten times more than its predecessors, but some estimates see this growing to two to four MB of memory.

If the meter is to become the communication hub, it would likely require an applications processor that would have a more sophisticated operating system, such as Linux, and also house the many different HAN communication stacks.

If the metering market were to evolve in this manner, ARM is well positioned to support that evolution. ARM has the Cortex Mx CPUs that are ideal for the real-time measurement and security functions of the meter. ARM can also scale all the way up to a Cortex Ax CPU, which has the Memory Management Unit (MMU) required to support a full featured operating system, such as Linux, and manage the multitude of communication protocol stacks.

In order to make the smart grid a reality the end user needs to be engaged. This means a whole new generation of products needs to be created that is both more energy efficient and increases energy consciousness. The ‘Internet of Things’, utilizing universal connectivity, cloud computing and visual computing, can incorporate energy awareness as part of the average consumer’s connected life.

The consumer wants to embrace energy conservation as well as the inherent personal cost savings. To achieve this we need those GUI interfaces to make managing the multitude of electrical devices seamless and painless for consumers, and to make end user products more and more intelligent and interactive.

What does the future hold?

‘Supersmart’ meters will be similar in functionality to the PCs we use today in that they will provide plug-and-play functionality. Electric metering is easier to accomplish than other utility monitoring as it parasitically draws energy from the power line it is measuring. By contrast, gas and water meters present a different set of technology challenges.

These meters require ultra-low power operation, as they are battery operated. The challenge centers on enabling these meters to communicate with the utility office cost effectively. One concept is to use existing networks, such as cable, cellphone networks or even the electric meter infrastructure as utility data portals.

This means a meter from one manufacturer may need to ‘talk’ to one from a different manufacturer. Such compatibility between meters is a significant advance in the meter technology being deployed today.

All meters, whether gas, water or electric, need to be able to communicate with one another regardless of the manufacturer. Meeting these compatibility requirements will be achieved with a set of standards that are still evolving.

Standards enable emerging markets to develop, similar to the way cable modems were first introduced to the computer market. When broadband cable was first being rolled out, the network providers would insist on a particular modem from a particular manufacturer.

Today, a consumer can go to any electronics store and buy a Data Over Cable Service Interface Specification (DOCSIS) modem, plug it in and it will work. In the future, will the consumer actually buy the meter at a local hardware store or an electronics retailer?


As mentioned above, a critical element of future meters is two-way communication capability. This capability will give rise to enhanced functionality. For example, the ability to load-balance energy across an entire nation will be possible.

Smart metering can incorporate energy awareness as part of the average consumer’s connected life Source: ARM

Imagine how significant the three-hour time difference could be within the United States. When the East Coast rises in the morning and energy demand goes up as people begin their day, a proportion of the associated energy demand can be provided by West Coast states, where everyone is still sleeping and demand is at a minimum. Conversely, when the East Coast winds down their surplus energy can be sold to the West Coast.

Energy trading can be filtered down to the individual consumer. If an individual decides to invest in personal energy creation technologies, such as solar or wind technology, there is the option to sell their excess energy to the smart grid.

Another example is with an electric vehicle’s battery array. If this battery has an excess charge then the owner might sell its reserves, only to recharge at a cheaper rate later while they are asleep. Energy can become a new form of currency that is enabled by smart meters and a smart grid.

In order to allow such activities, CPUs need to be able to provide the capability to complete secure transitions in real time. Such transitions must also be kept as simple and easy to use as possible to ensure adoption by the average consumer. The use of a ‘supersmart meter’ in this fashion should be no more complex than buying or selling an item online.

In addition to the base metrology function, the theorized ‘supersmart meter and grid’ now incorporates plug-and-play capability, meter-to-meter compatibility, multiple communication protocols, complex user interfaces, operating systems, field reprogrammability and secure transactions. When meters with this level of technological sophistication are in place, the potential benefits of smart metering and the smart grid will be achievable.

The smart grid will bring about a more energy conscious mindset amongst consumers. Smart meters as they are today are just the starting point. The smart grid will act as the backbone to allow data and two-way communication between end users and energy companies.

Consumer products present in the home, such as thermostats and other appliances, need to link and communicate with the meter and provide the data and information on energy utilization in a simple, interactive graphical interface. This will inform, engage and motivate the consumer to embrace energy conservation as a way of life.

These smarter meters and smarter consumer products will rely on an increase in software sophistication, which means more memory and more CPU bandwidth, fueling the need for 32-bit controllers to become ubiquitous.

Only when this ubiquity happens will the infrastructure and technology be in place for a truly Smart Grid, and the widely touted benefits of smart metering, for the consumer and energy companies alike, will be realized.

More Power Engineering International Issue Articles
Power Engineering International Archives
View Power Generation Articles on

No posts to display