The business world seems to be all
about “smart” these days. You
have smart cars, smart money,
even smart mobs. The utility industry
is no laggard in this wave of smartness.
“Smart meters” is one of the current
buzz phrases that quickly invoke discussions
about “smart grids” and eventually
“smart customers.” All this is via the
smart technology known as advanced
metering infrastructure (AMI).

But smartness will not come cheap.
The investment in AMI is significant;
north of $130 per customer at current
market costs. The business cases for
such projects focus heavily on the meter
– what type with what functions; the
network required to collect all the data
and what systems; new and current ones
modified that are required to process the
tidal wave of data AMI will generate. The
benefits are no more meter readers, far
more efficient operations and customers
who demand responsive.

Today it is estimated that only 10 percent
of U.S. customers have some form
of automated meter reading device. And
although that number will change by a
healthy level in the next five to seven
years, widespread implementation of AMI
is still in the planning and pilot stages at
many utilities. This is not out of desire to
implement AMI but because the business
case does not balance.

However, the industry has an opportunity
to change the balance of cost-tobenefit
in their business cases. By aggressively
adopting open standards, the cost
of AMI meters can be reduced significantly
and the applications that “harden” some
of the “softer” benefits in business cases
can be effectively developed. The fundamental
shift in thinking that is required is
the realization that AMI is not a meter system
replacement. It is transformational. It
is a classical innovators’ dilemma. AMI is a
computer and communications system. It
is hardware and software. Forget meters.
Think computers. Think applications!

Let’s first take a look at the key components
of AMI – the smart meter, the
smart customer and the smart utility
– and then consider the challenges this
vision presents.

Smart Meter

AMI starts with a smart meter. The frequent
question is, how smart? Is an AMR
meter with some application integration
in the back end sufficient?

AMR automates the collection of meter
reads. No more meter readers doing their
appointed rounds on a monthly cycle.
Instead trucks make the rounds at up
to 35 miles per hour and AMR meters
transmit their reads. Or a fixed network of
varied technologies collect the reads, perhaps
more frequently than monthly, but
all to the same end – replace people with
automated meters.

But consider the AMI vision. Your
electric meter is really “smart”; it has an
embedded computer; it tracks and stores
all the vital information about your energy
use; and it communicates your usage and
other interesting information back to the
utility on demand. It will send out a distress
call to the utility when the power is
out. It can connect to an in-home, IP-based
LAN that communicates with all your
appliances, monitoring them and, in times
of power shortages, can even change the
operations of these appliances.

AMI requires a meter that is functionrich
and able to provide information in
near-real time. It needs to be two-way,
upgradeable, programmable and extensible.
This ups the ante when the decision
as to what type of meter and how smart it
needs to be is considered. Accordingly, it
significantly increases costs.

Most business cases treat the AMI
meter as 20-year utility property. With
over 75 percent of the AMI cost in the
meter, its associated network and its
installation cost, changing out this infrastructure
in a near- or medium-time
horizon will be a financial problem to the
business case. A five-year horizon like
that used for computer equipment would
quickly sink most business cases.

Nonetheless, conceptualizing the meter
as a computer versus a utility meter is the
defi ning change in perspective required
for achieving AMI benefi ts (see Figure 1).
And with this realization enters the opportunity
for the fundamental, aforementioned
shift. The AMI meter is a computing
platform and thus the value comes from
the applications it supports – not just the
functions it replaces or automates. These
are applications that support the smart
customer and the smart utility.

The Smart Customer

For AMI benefi ts to be realized you need
a “smart customer.” A smart customer is
both informed and empowered in regard
to their energy use; informed because
they know their energy use and its associated
cost over time and how that energy
use is derived from all the energy-consuming
devices in their premise.

Moreover, a smart customer is an
empowered customer. It does little good
for a customer to receive hourly or even
sub-hourly energy and cost information
if they have no idea what to do with the
information. They need access to a set
of easy-to-use applications that allow
them to visualize the cost of their energy
behaviors and to make informed decisions
on how to change those behaviors.
These applications need to have the ability
to perform what-if analyses to support
customers’ energy and cost-effi cient
behaviors. And even more ambitiously,
the AMI system needs to deliver real-time
price signals that alert customers of highcost
conditions and that allow utilities
to send signals directly via a meter connected
to a customer-premise LAN that is
able to control customer energy-consuming
devices.

The Smart Grid

Utilities need to be smart for the business
vision to work.

Utility support for operations in today’s
world is basically forensic in nature. Take
outage management, for example. Generally
the utility waits until enough customers
call to report an outage to triangulate
on the source of the problem and begin
restoration operations. In the case of a
grid component failure such as a transformer
explosion, they wait until the
explosion happens, usually unaware of
its impending failure. Why? Because they
have generally no telemetry installed
to give them the required information
before the customer calls or the transformer
blows.

But with AMI meters deployed to all
customers, the utility gains the ability
to continuously monitor the operational
status of the network at the end points.
Network component failures are instantaneously
observable and can be isolated
directly to the limited portion of the grid
impacted. During an outage, the utility
already knows a customer’s power is out
and is able to provide a timely and accurate
estimate for the time of restoration
when a customer calls.

On an ongoing basis, the history of
key metrics such as distribution loading
– from which key operational metrics of
transformer performance can be induced
– are collected and analyzed. Predictive
maintenance and the replacement and
upgrade of network devices are managed
more effectively and effi ciently. The
distribution grid is managed predictably
instead of forensically because the AMI
system can operate as a proxy for many
of the historical functions of OMS and
SCADA.

Challenges

If this compelling vision of the benefi ts of
AMI did not have its challenges, the current
less than 10 percent penetration rate
of AMI in the U.S. would be much higher.
It turns out the challenges are formidable.

The first challenge is the cost for the
function required. AMI meters are very
expensive relative to their predecessor, the
electromechanical meter. Clearly, a meter
that is much “smarter” and more function-
rich than an AMR meter is required
to realize the benefi ts of AMI. Such AMI
meters exist in the market today and are
generally too expensive to be deployed
on a large scale to all customers. A signifi –
cant element of the cost is in its research
and development. As R&D is amortized
and manufacturing capabilities grow and
mature, one can expect the cost to decline,
which has already occurred over the past
few years. But for AMI meters to reach
the near commodity status of today’s
electro mechanical meters, the industry
needs take a page out of the computer and
telecommunications industries’ playbook.
They need to embrace standardization
and openness in this technology. AMI is,
after all, a computer and communications
network, and the recent history in both
industries is a dramatic example of what
open standardization can do to drive down
costs and deepen functionality.

The second challenge focuses on the
set of applications required to realize
both the smart customer and the smart
utility. These are customer applications
that can provide the information that customers
will access to become informed
and empowered. These are also utility
applications and analytics that support
operations, as well as the integration of
these applications into the operational
processes of the utility. Both sets of
applications require a fundamentally different
view within the utility in regard to
the basic systems and business process
supporting customer information and
utility operations.

Implementing these systems both
through new systems and modification
and adaptation of current systems is
complex. Most business case analysis
focuses on the obvious issue of scalability
due to the large increase in raw data
processed within the utilities systems.
But the problem is much more pervasive
than data volume. The integration
required between batch-orientated billing
systems, real-time outage systems and
near-real time customer information analytics
applications – all utilizing the same
data stream from the same AMI system
infrastructure – is as complex and sophisticated
as any system the industry has
yet implemented. The important starting
precept is that the AMI system is a complex
collection of multiple, interrelated
computer and communications systems,
not all of which are operating at the same
temporal cadence. Effective integration
of these systems is the controlling problem
impacting successful implementation.

The process of planning to add a new
meter data management system that connects
to the AMI network, and connecting
it to the current utility systems through
currently employed techniques, will more
than likely deliver sub-optimum results.
AMI systems cannot be glued together
in some gerrymandered architecture. A
much more robust, flexible and extensible
architecture is required. Again, as with
the meter, standardization and openness
is an absolute requirement. Building
systems around Web services and
SOA architectures is already adopted by
industries with the same fundamental
problem of building similarly complex and
extensible systems. Only through such
open and standard state-of-the-art architectural
approaches will AMI applications
be successfully and effectively integrated
to perform all the functions required to
achieve AMI benefits.

Adopting open standards for both the
meter and customer and utility application
development will drive the cost of
AMI implementation down dramatically
and increase the realizable benefits to
both the customer and the utility. It will
also enable the level of application innovation
required to deliver the benefits
of AMI. AMI champions need to focus on
computer systems and applications – and
not on the replacement of electrical measuring
devices – to be successful.

It’s the smart thing to do!