Social innovation enablers to unlock a low energy demand future

We initiate the process of developing a comprehensive low energy demand (LED) innovation narrative by applying the framework ‘Functions of Innovation Systems’ (FIS) and identifying the key conditions under which technology interventions can be improved and scaled up over the next three decades to contribute to climate change mitigation. Several studies have argued that the potential for LED-focused mitigation is much larger than previously portrayed and have shown that adopting a wide variety of energy-reducing activities would achieve emissions reductions compatible with a 1.5 C temperature target. Yet, how realistic achieving such a scenario might be or what processes would need to be in place to create a pathway to a LED outcome in mid-century, remain overlooked. This study contributes to understanding LED’s mitigation potential by outlining narratives of LED innovation in three end-use sectors: industry, transport, and buildings. Our analysis relies on the FIS approach to assess three innovations in these sectors. A key insight is that the distinct characteristics of LED technology make enabling social innovations crucial for their widespread adoption. Finally, we identify a set of eight social enablers required for unlocking LED pathways.


Introduction
The low energy demand (LED) scenarios (Grubler et al 2018) quantitatively showed that LED technologies and behaviors could reduce emissions over the next 30 years compatible with a 1.5 C target.They used detailed estimates of how low energy use could become across the whole economy and then used integrated assessment modeling (IAMs) to evaluate mid-century impacts on greenhouse gas emissions and other social indicators.It built on earlier efforts going back to the 1970s, for example Soft Energy Paths (Lovins 1976), as well as the International Energy Agency's longstanding work on energy efficiency (IEA 2022), but is distinct because of its low temperature target.The LED results are important because they showed that the scope of emissions savings through lower energy use is vast, to the extent that they enabled the Paris Agreement targets to be achieved without relying on technological carbon removal.These insights arose from the authors' distinct focus on energy services and whence energy demand arises-particularly in the context of increasing access to energy services in developing countries.
The original LED paper had a substantial impact on the climate community.For example, its insights were prevalent in the IPCC AR6 chapter on energy demand (Creutzig et al 2022b).However, one prominent critique is that while the paper calculated careful estimates of energy savings, it did not make claims about how realistic achieving those savings might be (Keyßer and Lenzen 2021).While (Grubler et al 2018) compared energy use in today's world to that in the LED future through IAMs scenario runs, it did not explicitly characterize how behavioral and social dynamics would enable the transition to a low-energy world (Cordroch et al 2022).For example, it provided only an overarching view of how digital technologies may impact energy demand from industry and adopted a quite specific focus on energy savings due to the increasing role of services in meeting consumer choices.This set of mechanisms is particularly important given the LED scenario's emphasis on energy end-use, the role of consumers, and especially how they interact with each other.The extent to which LED approaches are taken seriously as credible complements to other mitigation efforts depends on understanding whether a LED world can be realistically attained over the next three decades; this in turn depends on a better characterization of the mechanisms for a transition to an LED world, which we describe here as LED innovation.
In this paper, we initiate the process of building a comprehensive narrative on LED innovation that describes how LED technologies can improve and be scaled up to become widely adopted.Following (Nemet and Greene 2022), LED innovations are defined as 'any effort to improve the level and structure of energy demand' , 'including LED technologies and LED services.'This broad definition allows us to depart from a purely technological focus towards the inclusion of human behavior and business models.We rely on the 'Functions of Innovation Systems' (FIS) framework (Bergek et al 2008) to identify innovation dynamics and adoption timelines consistent with overall LED scenarios.This allows us to identify conditions and key enablers for the successful development and deployment of LED innovation in the context of five key drivers of change that condition innovation in LED (Grubler et al 2018): higher living standards, urbanization, digitalization, novel services, and prosumers.This paper thus represents a first step towards the generation of narratives characterized by a comparative systemic perspective through both quantitative indicators and qualitative descriptors to appropriately capture the many ways LED futures could unfold, given the distinct characteristics of LED innovations (Nemet and Greene 2022).

Analytical approach
Following Grubler et al (2018), we focus on three key high-level sectors-buildings, transport, and industry-because they provide a useful taxonomy congruent with the structure of IAMs.Essential improvements in terms of LED and decarbonization also arise when these sectors are coupled, such as vehicle-to-grid applications (Noel et al 2021).In addition, general purpose technologies with pervasive effects across sectors also play a key role, e.g.digitalization (Wilson et al 2020b, Creutzig et al 2022a).
Our comparative case study analysis of innovations relevant in the context of a LED 5 follows the approach of (Hekkert et al 2007), which identifies seven key FISs: knowledge development, knowledge diffusion, the guidance of search, resource mobilization, entrepreneurial activities, market formation and creation of legitimacy.We conceptualize how the development from nascent adoption to widespread diffusion of three innovations-which differ in innovation stage and maturity-may support the achievement of LED with a specific focus on the relevance of the different functions, and the role that key actors will play in LED scenarios in various sectors.While other innovations are also relevant in the context of LED, the three examples we discuss here are both relevant for their impact on energy demand, and widely discussed in both academic and business circles.A main insight from this comparative approach is that successful LED innovations involve substantial contributions from, and interactions among, technological, social, and business model innovation.This ultimately leads us to identify and discuss key enablers and conditions for further development and widespread uptake relevant for modelling LED pathways.

Distinct characteristics of LED innovations
A set of characteristics makes LED innovations distinct from many other types of innovation (Nemet and Greene 2022).In our context, these can be summarized as follows: 1. LED innovations favor a movement from a traditional economy based on goods towards the provision of Services.However, more efficient service provision can stimulate more utilization through short-term and longer rebound effects.

Case study selection
Among the many possible cases, we selected one for each sector, which has potential to contribute to LED.
Table 1 provides an overview of the main characteristics of the three case studies: additive manufacturing (AM), which has the potential to revolutionize the production of goods within the industrial sector; sharing mobility (SM) as a cutting-edge service within the transportation sector; and household energy users producing electricity with rooftop solar to become solar prosumers (SP) in the realm of building infrastructure.These cases exhibit distinct characteristics, including the nature of the services, adopters, and technologies.They also represent varying stages of innovation, while all three hold significant promise for reducing energy demand.Another vital factor guiding our selection process was the need to illustrate the diverse roles of enabling conditions in expediting adoption by end-users.

Three LED innovation case studies
For each of the three case studies, we summarize the context within which the innovation is emerging, its potential demand reduction, and the health of its innovation system functions.For more on the case study FIS analysis, please see SI1-3.

Unlocking LED technological innovations pathways
To maximize their potential, LED innovation necessitates coordinated efforts from individuals, businesses, cities, regions and countries (also see SI. Importantly, investment patterns will change.For example, the insurance industry might be affected by its role as an insurer of third-party liability.

LED narratives
We offer illustrative LED narratives for the case studies of AM, SM and SP to highlight the crucial roles that the eight enablers can play in optimizing the potential of LED technological innovation (see figure 1).Because of the distinct characteristics of these cases, in addition to context and scale, all or only a few of these enablers may be applicable.Even though each case is at a different level of maturity, they are all at an early stage in their development relative to the level required to achieve 1.5 C. AM LED narrative.By 2050 it is plausible that AM technologies will have significantly increased their penetration rate, but their potential to contribute to LED varies by sector: AM could reduce total primary energy supply in aerospace fuels by between 9 and 35%, from aerospace manufacturing by between 8 and 19% and in the sectors of medical equipment and tools by between 5 to 19%, and 3%-10%, respectively (Gebler et al 2014).Three critical junctures exist in the AM narratives.A first is whether the technology will mature and proof-of-concept will be achieved for large-scale manufacturing.Achievement of proof-of-concept in several key areas-such as metal-based AM or AM based on recycled inputshinges on public support for innovation in the form of technology-push policies, including R&D investments and subsidies.Application of AM technologies in other sectors and their relevance for decarbonization is dependent on lifestyle changes and developing a culture of sufficiency which would avoid material and energy rebound.A second is whether new business models will be fostered to move away from traditional linear manufacturing and towards on-demand production.Network effects and changes in business models are necessary to promote the scale-up and give rise to learning-by-using dynamics, including increased technology efficiency and lower costs.Access to finance and investment needs to be granted to promote adoption among producers and service providers.A third is whether governance will promote the necessary behavioral change, mindsets, and targeted policies to ensure AM achieves deep emission reductions, circularity, and sufficiency as well as economic benefits.Tailored public policies, including regulation and the development of industry standards, are needed to promote a paradigm shift from linear and subtractive manufacturing to AM consistent with a deep sustainability transition.
SM LED narrative.Shared mobility innovation demonstrates various levels of development and adoption linked to the economic, social, institutional, and financial contexts.For example, traditional micro-mobility like cycle rickshaws thrive in developing countries, while bike-and scooter-sharing systems flourish in developed countries.Drawing from the literature, several social enablers are relevant and have the potential to significantly influence the transition toward realizing the full LED innovation potential.Specifically, these enablers include design and access to technology and infrastructure, exemplified by carsharing's global expansion and shared electric vehicles' emission reductions (ITF 2017(ITF , 2019(ITF , 2020)).Well-designed, safe and accessible infrastructure provides the necessary support to be effectively utilized and integrated, encouraging individuals and organizations to embrace and benefit from the advancements (Goodman et al 2014, Song et al 2017, Creutzig et al 2022b).Policy packages such as social and behavioral intervention, like nudges, can influence individuals' choices in favor of SM usage, e.g.'walk cycle ride' campaign (Rojas López and Wong 2017), while carbon pricing mechanisms can affect incentives (Eliasson andMattsson 2006, Richardson et al 2010).Social movements and peer effects exist in that individuals are often motivated by collective sentiments and the behaviors of their peers when embracing new services and lifestyles (Burghard andDütschke 2019, Whittle et al 2019).Another enabler, business models, includes expanding service offerings, fostering collaborative partnerships, and optimizing pricing, as well as providing adequate finance and investment support to ensures the development and expansion of SM infrastructure and services, making them more accessible and attractive to the public.
SP LED narrative.Rooftop solar is mature and growing rapidly, but at less than 2% of global electricity supply, its future potential remains large, and the extent to which that potential is realized depends highly on social innovations and behavioral change in particular.Two key junctures in the rooftop PV narrative are 1) whether growth in adoption continues in existing markets, especially in East Asia, Europe, and North America, and 2) whether adoption proliferates in emerging low-and middle-income countries.If both occur, we will see rooftop PV take the upper pathway to reach well above 10% of the global electricity supply, perhaps accounting for half of all PV, which could account for half of electricity supply.That pathway depends on key social innovations.Existing market growth depends on governance and regulation, particularly supportive electric rate structures, credit for energy storage, and incentives for reducing demand and would be bolstered by the proliferation of peer-to-peer electricity trading, which itself requires modification of infrastructure design.The latter would give consumers more control and would take full advantage of the granularity that rooftop solar provides; peer effects combined with network effects could catalyze further adoption of rooftop solar.Emerging market growth depends on all the above, as well as a fundamental change in finance and investment, particularly credit access, which is central to improving costs and market development.Financing for rooftop solar remains much more expensive than in existing markets, which in combination with lower median incomes, limits adoption.Business model innovations that can provide credit at rates like those in existing markets would help put rooftop solar on the upper pathway.One can also imagine a scenario where none of these supporting social innovations take hold, and adoption grows at a rate that only grows with overall demand leaving its share unchanged.While many middle pathways exist, an especially problematic one has supporting social innovations take hold in both existing and emerging markets, but access to finance remains a constraint in emerging markets.That would lead to a tremendous, wasted opportunity both for the global impact and distribution of the access to benefits that rooftop solar can provide.

Conclusion and directions for further work
This paper examines three pertinent innovations in LED through the lens of the FIS framework: AM, shared mobility, and SP.Furthermore, it discusses key enablers to unlock their LED potential and demonstrates ways for scaling them up to widespread adoption.We highlight that technological innovation is not the sole driver of LED.Rather, realistically achieving LED requires a range of crucial nontechnological, non-cost aspects of technology diffusion, such as changing behaviors, social norms, and governance, all in the context of heterogeneous agents and the development of local knowledge.Consequently, this study underscores the need for additional bottom-up qualitative and quantitative research and analysis.This approach can offer a more in-depth understanding of the complexities and, potentially, serve as a valuable complement to addressing the challenges posed by large-scale IAMs.
To present feasible LED scenarios, capture key enablers and explore (non-linear) solutions, current scenario modelling needs three key shifts.First, largescale IAMs and energy system models could be complemented with modelling approaches which can more realistically capture heterogeneous effects and the role of peer effects in technology development and diffusion.This is the case, for instance, for agentbased modelling, which has the potential to play an increasing role in informing large-scale modelling exercises (Niamir et al 2020a, Edelenbosch et al 2022).Second, the key role of non-technological drivers in LED innovation points to the importance to develop more detailed LED narratives to ground and justify model results-that is, provide rationales for specific model constraints to account for changes in lifestyle, norms, and beliefs.Finally, available IAMs are extremely limited in their ability to depict different policy instruments and mixes.This surely represents an interesting area of further model development and is a particularly important one in the context of LED pathways.
To improve our understanding of LED innovation and narratives alongside our modelling tools, there are two complementary fruitful avenues of future work.On the one hand, detailed empirical data and analysis on LED innovation development and adoption, particularly in LMI countries, is needed to characterize the intersections between institutional, political, environmental, social, and economic factors which contribute to LED innovation development and adoption at different scales in different geographies (Beckage et al 2020, Rubiano Rivadeneira andCarton 2022).On the other hand, future research should inform and complement IAMs by distilling specific insights regarding policy instrument and mixes in support of LED innovation.This includes valuable insights on establishing and promoting policy motivations, addressing innovation system failures, improving technologies and services, facilitating behavioral change and LED innovation adoption, enabling new business models, and addressing adverse consequences of successful adoption-e.g., material or energy rebound or distributional repercussions.Attention should be given to the role of local context both in adoption and policy, especially in LMI countries and to the combination, timing and strategic sequencing of policy instruments.Pursuing this comprehensive research agenda can help unlock the potential of LED scenarios in the context of urgent deep decarbonization.

Figure 1 .
Figure 1.Illustrative LED innovation pathways.Lines indicate the cumulative adoption of an LED innovation over time, with key junctures indicated by knots.Diverging LED innovation pathways (orange to blue) illustrate that interacting choices and actions by diverse actors-government, business, and civil society-can enable LED innovation adoption.Pathways and opportunities for action are shaped by the innovation stage, distinct characteristics, and previous actions (or inactions and opportunities missed) and enabling and constraining conditions.Social innovations can create enabling conditions allowing for steeper adoption pathways (blue pathways).At the same time, lack of knowledge, finance and investment, institutional and social drivers as well as poverty and inequity, halt the process (orange pathways).{The concept of the visualization of illustrative pathways is derived from figure SPM.6 (IPCC 2023)}.

Table 1 .
Summary characteristics of the case studies.
Yet, typically, rebound takes back only some of the energy savings but not all(Gillingham et al 2016).tion expands the system boundaries of our analysis: from a purely technological focus towards the inclusion of human behavior and business models; from a pure focus on energy to consideration of material use impacts; and from a mere focus on monetary indicators (e.g.GDP) towards inclusive well-being.Our working assumption is that successful LED innovations are likely to involve substantial contributions from, and interactions among, technological innovation, social innovation, and business model innovation.
(Denholm and Margolis 2016) economies of scale, indicating limits to large-scale diffusion (Sculpteo 2014, Baumers et al 2017, Steenhuis and Pretorius 2017).7.Creation of legitimacy: governments and the education sector generate legitimacy for AM technologies (e.g.inclusion of AM-relevant and specific training as part of undergraduate and graduate curricula) (Reis 2013).Yet, large-scale adoption of AM is inhibited by technical hurdles and lack of universal guidelines for metrology, inspection, and standardization, leading to concerns about IP protection and vulnerability to cyberattacks.Technologies enabling SP are mature (O'Shaughnessy et al 2018a), the challenge being their integration in the grid, a form of infrastructure design.Therefore, social innovations and behavioral changes, especially peer effects, have a large role to play.The proliferation of household energy storage, either standalone or in electric vehicles, combined with digitalization, especially connectivity to the grid, would facilitate a larger role for SP(Denholm and Margolis 2016).Knowledge diffusion: components and know-how are widely available and transferring grid integration experience from areas with high solar adoption is now a focus (Heptonstall and Gross 2021).3. Guidance of search: public policy such as renewables obligations, subsidies, as well as information programs (like Solsmart in the US) raise awareness and orient expectations of growth.Solar is very popular (Roddis et al 2019, Hazboun and political forces use extant regulations and natural monopolies to bar the inception of neighbors buying and selling electricity. (Aboulkhair et al 2019)etrically complex structures in a single-step process (Reis 2013); (2) major cost advantages in sectors where products need to be adapted to customers' needs(Aboulkhair et al 2019).1. Knowledge development: maturity of AM varies greatly by sectors; in no sector is it applied for large-scale manufacturing (AMFG 2019).Universities are a main source of codified knowledge, but patenting is dominated by the private sector (Peña et al 2014).Innovation in new materials and methods for AM is progressing (Ngo et al 2018).2. Knowledge diffusion: collaboration among different actors required for large-scale application of AM (Lavoie and Addis 2018).Ngo et al (2018).3. Guidance of search: many governments consider AM a technology of interest for industrial policy, but their strategies differ significantly, ranging from targeted public R&D funding (Peña et al 2014, McKinsey 2017, Samford et al 2017) to fostering synergies with existing local industrial strengths (Peña et al 2014, McKinsey 2017, Samford et al 2017).4. Resource mobilization: early-stage financing for AM innovation and funds for further development came both from private and public sources.(Peña et al 2014), including the automotive 6.Market formation: in 2017 the market for AM products and services was over $7 billion.(Thompson et al 2016).Yet, the penetration of AM technologies is estimated at only 8%, indicating untapped potential (Vora and Sanyal 2020).3.2.Case 2. Shared mobility (SM) SM, characterized by asset sharing (e.g. a bicycle, escooter, vehicle) and facilitated by information technology (e.g.apps and the internet), holds promise for emission reduction and climate change mitigation (Creutzig et al 2022b).Four business models have been identified (Santos et al 2018): peer-topeer platform (Ballús-Armet et al 2014); short-term rental managed and owned by a provider (Enoch and Taylor 2006, Bardhi and Eckhardt 2012, Schaefers et al 2016); Uber-like service (Wallsten 2015); and shared ride where private vehicles shared by passengers to a common destination (Liyanage et al 2019, Shaheen and Cohen 2019).Studies show that SM, mainly shared automated electric vehicles, tionality, and future iterations of shared mobility solutions (Ruhrort 2020).This iterative learning process, often referred to as 'learning by using,' nurtures innovation by incorporating direct user feedback into the R&D cycle (Hekkert et al 2007).Consequently, the network's role transcends traditional knowledge exchange to become (Denholm and Margolis 2008, Michaels and Parag 2016, Gagnon et al 2018) and the global potential is vast (Creutzig et al 2017, Haegel et al 2019).The global mitigation potential of solar energy is 2-7 GT CO 2 /year in 2030, among the highest of all mitigation options (Babiker et al 2022), with rooftops providing a substantial share.the policy regime (Horstink et al 2021) although in some jurisdictions growth has stalled or has been thwarted by opposing actors.An inclusive governance approach to regulation and policies would help establish legitimacy.Peer-to-peer energy trading, the logical next step for prosumers (Luo et al 2014), has not yet achieved legitimacy and strong agent-based technology choices in an integrated modeling framework iScience 25 103905 Egli F, Polzin F, Sanders M, Schmidt T, Serebriakova A and Steffen B 2022 Financing the energy transition: four insights and avenues for future research Environ.Res.Lett.17 051003 Eliasson J and Mattsson L-G 2006 Equity effects of congestion pricing: quantitative methodology and a case study for Summary for policymakers Climate Change 2023: Synthesis Report.Contribution of Working Groups I, II and III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change ed Core Writing Team, H Lee and J Romero (IPCC) pp 1-34 ITF 2016 Shared Mobility: Innovation for Liveable Cities (available at: www.itf-oecd.org/shared-mobility-innovation-liveablecities)ITF 2017 Transition to shared mobility: how large cities can deliver inclusive transport services (ITF (available at: www.itf-oecd.org/transition-shared-mobility)ITF 2019 ITF Transport Outlook 2019 (OECD Publishing) (available at: www.oecd-ilibrary.org/transport/itf-transportoutlook-2019_transp_outlook-en-2019-en)ITF 2020 Good to go? Assessing the enviromental performance of new mobility (OECD/ITF) (available at: www.itf-oecd.org/good-go-assessing-environmental-performance-newmobility) Keyßer L T and Lenzen M 2021 1.5 • C degrowth scenarios suggest the need for new mitigation pathways Nat.Commun.12 2676 Lavoie M and Addis J L 2018 Harnessing the potential of additive manufacturing technologies: challenges and opportunities for entrepreneurial strategies Int.J. Innov.Stud. 2 123-36 Lee H and Song H J 2021 Current status and perspective of colored photovoltaic modules Wiley Interdiscip.Rev. Energy Environ. 10 e403 Li Z 2023 Leveraging explainable artificial intelligence and big trip data to understand factors influencing willingness to ridesharing Travel Behav.Soc. 31 284-94 Liyanage S, Dia H, Abduljabbar R, Bagloee S, Liyanage S, Dia H, Abduljabbar R and Bagloee S A 2019 Flexible mobility on-demand: an environmental scan Sustainability 11 1262 Lovins A B 1976 Energy strategy: the road not taken?Foreign Aff.55 65-96 Luo Y, Itaya S, Nakamura S and Davis P 2014 Autonomous cooperative energy trading between prosumers for microgrid systems.39th Annual IEEE Conf. on Local Computer Networks Workshops pp 693-6 Ma Y, Rong K, Mangalagiu D, Thornton T F and Zhu D 2018 Co-evolution between urban sustainability and business ecosystem innovation: evidence from the sharing mobility sector in Shanghai J. Clean.Prod.188 942-53 Markard J 2018 The next phase of the energy transition and its implications for research and policy Nat.Energy 3 628-33 Martinez L M and Viegas J M 2017 Assessing the impacts of deploying a shared self-driving urban mobility system: an agent-based model applied to the city of Lisbon, Portugal Int.J. Transp.Sci.Technol.6 13-27 McKinsey 2017 The great remake: manufacturing for modern times (McKinsey) (available at: www.mckinsey.com/capabilities/operations/our-insights/the-great-remakemanufacturing-for-modern-times) Michaels L and Parag Y 2016 Motivations and barriers to integrating 'prosuming' services into the future decentralized electricity grid: findings from Israel Energy Res. Soc.Sci.21 70-83 Mouratidis K 2022 Bike-sharing, car-sharing, e-scooters, and Uber: who are the shared mobility users and where do they live?Sustain.Cities Soc.86 104161 Neij L, Heiskanen E and Strupeit L 2017 The deployment of new energy technologies and the need for local learning Energy Policy 101 274-83 Neij L and Nemet G 2022 Accelerating the low-carbon transition will require policy to enhance local learning Energy Policy 167 113043 Nemet G and Greene J 2022 Innovation in low-energy demand and its implications for policy Oxford Open Energy 1 oiac003 Ngo T D, Kashani A, Imbalzano G, Nguyen K T Q and Hui D 2018 Additive manufacturing (3D printing): a review of materials, methods, applications and challenges Composites B 143 172-96 Niamir L, Ivanova O and Filatova T 2020a Economy-wide impacts of behavioral climate change mitigation: linking agent-based