How IT Can Increase Demand Response Participation
The Role of AI and IT Systems in Scaling Residential and C&I Demand Response Programs
Electrical utilities are, as the expression goes, in a “real pickle.” The challenges they face are
myriad–ranging from meeting increased load demand from data centers to onboarding dynamic
generation from renewable sources. Still, demand response (DR) continues to be a strategic
part of the utility coping tool box.
DR is now a part of a utility’s grid flexibility strategy. However, especially amongst residential customers, customer participation continues to be a vexing issue. For example, a 2025 report indicated that only 20% of households with a smart thermostat participated in a DR program. Increasing customer participation is now a high priority item. While DR traditionally has been
more of an analog process, it is rapidly being digitized. Increasing adoption of information
technology (IT) based systems, both on the utility and the customer side, holds great promise to
increase participation rates. DR can be expected to be fully automated in the near future as it
becomes a feature of IT-based systems such as virtual power plant (VPPs) and Energy-as-a-Service (EaaS). With AI being the core of these systems, the current AI boom will only accelerate their capabilities and adoption.
Residential DR: the toughest nut to crack
As many residential customers already own smart thermostats, making it easier to participate in
DR programs, theoretically residential DR should be booming. However, the perception
amongst utilities is that these programs are unreliable since it’s easy for customers to just adjust
their temperature settings for increased perceived comfort.
With ever-more-capable AI making it simpler to parse and use multiple data sources, residential
DR could be improved. One could imagine a use case where utilities would use an AI agent to
message their customers with information on what their potential savings and payments could
be if they keep their temperature settings as is.
The adoption of AI-powered building energy management systems (BEMS) in multi-family
residential units is a bright spot. Typically using AI-based analytics of sensor and other
energy-related data, BEMS can provide building owners with insights into and optimization of
energy usage in their buildings and even the ability to control devices. BEMS can allow building
owners the ability to participate in DR programs and leverage their capabilities to gain the
maximum benefit from their participation.
C&I DR: A brighter spot?
The benefits of DR programs are more lucrative for commercial and industrial (C&I) customers,
however they come with a number of customer concerns. These can range from concerns about
the impact on production schedules to whether a stopped system will work properly upon
restart. Here too, AI-enabled IT technologies hold the possibility of increasing DR participation.
One is predictive maintenance systems. These systems use AI to analyze sensor data streams,
such as vibration, to find patterns which predict when a system is likely to need servicing to
avoid downtime. For example a number of mining companies use predictive maintenance to
monitor the health of conveyor belts and other mining equipment. In a DR scenario, an utility
C&I customer using predictive maintenance can use it to identify equipment that can be shut
down for maintenance or benefit from down time during the DR time slot.
Another example is digital twins. In C&I settings, digital twins are software simulations of a
particular asset or even an entire building or campus. Using data streams from the actual
systems and sensors and, of course, AI, digital twins can simulate physical systems in near
real-time. An example here is BMW which maintains digital twin simulations of all 31 of its
production facilities. With a digital twin in place, operators can better identify systems in a facility
that are less likely to be impacted by DR participation or take on additional loads to cover for
shuttered systems.
Looking to a more automated future
Going forward, there are two technology trends that may ultimately subsume and automate DR
for both residential and C&I customers.
First is virtual power plants (VPPs). VPPs are essentially aggregations of resources such as
behind-the-meter batteries, hot water heaters, and HVAC systems managed by software
platforms in response to a grid request. Operated by either a third-party provider or a utility,
VPPs use AI to determine which assets it can call on, manipulate them, and, perhaps more
importantly, calculate payments made to the resource owner. For example, VPP could automate
a typical DR technique such as reducing water temperature of a home hot water heater since it
knows the heater is unlikely to be used for a while along with discharging energy stored in a
battery to the grid. In the summer of 2025, VPPs have already contributed to avoiding blackouts
in heatwaves.
The other is Energy-as-a-Service (EaaS). EaaS starts off as a business model doing exactly
what it says, providing energy to a customer as a service. Typically targeted at C&I customers,
EaaS can include a microgrid installation at the customer’s facilities and energy management
services among other offerings. One such solution is offered by Redaptive. Since DR represents
a financial opportunity for the EaaS provider, they could contract and manage DR for their
customer and reflect the savings in the energy price.
Each of these solutions recognizes that DR is just one of the flexibility tools that benefits both
the customer and the utility. With AI capabilities constantly expanding, the capacity for IT
systems to greatly automate DR and combine it with other services will only increase. Soon we
may be in a future where DR is largely automated and the customer rarely has to think about it.
This think piece was written by Matthew Tartaglia of Summit Human Capital on September 10th, 2025. Sources are cited within the article.