Dan and Anthony had a vision for the green home energy market – to align and automate different home energy devices to work efficiently as one, using a device that would seamlessly and securely integrate with the setup in people’s homes to make best use of them all.
Swarm’s concept would create a system that automated home energy decisions for users. Automation would transform how users manage, store and redistribute their energy. And a single device would make intelligent decisions on their behalf and save them money.
Taking an innovative technology product to market can be a challenging and time-consuming process, particularly for startups with a limited funding ‘runway’. Typically, it takes up to a year for a new product to hit the market. Swarm wanted to accelerate that process.
Green home technology is also relatively new, with little by way of an industry standard. Integrating sustainable devices would mean working with multiple individual circumstances. With Swarm developing a first-to-market product, it was also critically important that the new product and service would meet user needs.
Swarm’s technology would direct power based on people’s energy tariffs and usage, as well as the requirements of the national grid. It would introduce a tech-agnostic solution to an industry where there had previously been no open standard between solar manufacturers.
Swarm wanted to test hypotheses for its proposed new product. Could it create something that could communicate across multiple inverter types? Could it automate power consumption with users’ consent? And could it orchestrate multiple clients in different geolocations to do the same thing within a second of each other, confirming the feasibility of its ‘virtual power plant’ idea?
In late 2023, Swarm’s founders asked Opencast for help in turning their vision into a reality – and Opencast proposed an innovative and flexible partnership that would see it embedding a full-service team into the Swarm operation to develop its product in a live state, at scale. Opencast would share in Swarm’s future success through equity participation.
Collaboration would be easier with both firms headquartered at Hoults Yard in Newcastle. Demonstration and testing of Swarm’s technology could also take place at Hoults, in a purpose-built energy hub.
Opencast would help Swarm not just to create software, but also build a device and a service. This meant building a physical device with software, screens and designing an app, with cloud-based infrastructure to support the deployment, orchestration and monitoring of over 100,000 devices.
Opencast also helped Swarm to shape its company structure to enable rapid scaling up – as well as build the in-house tech team to ensure the business would be more self-sufficient.
With Opencast’s team exploring the technical and regulatory possibilities, as well as value and innovation, Swarm could turn its founders’ theory into a product that landed well in the market.
To identify customer needs, motivations and pain points, Opencast UCD consultants talked to home energy users. Insights from pilot home users helped to shape and inform Swarm’s minimum viable product (MVP).
Service designers would help deliver the entire service solution, including both front and back end. Interaction designers, meanwhile, would focus on the product’s front end.
The team’s product owner (PO) documented Swarm’s requirements, creating a product roadmap. They used a ‘wall of work’ to identify and document various work strands centrally. This would ensure that every team member understood how their work contributed to the product development overall.
Opencast’s developer team researched, defined and created a smart algorithm that effectively became Swarm’s AEM. The algorithm would decide every action the new product should perform. With imminent energy prices a known quantity, the algorithm would understand users’ needs, solar forecasts and energy tariffs, assessing when users should buy energy, when they would need it for consumption – and when to take advantage of high export prices by selling energy back to the grid.
The developer team, supported by a data scientist, used machine learning and deployed linear algorithms to address home energy import and export efficiency, handling and integrating complex datasets from a variety of sources, while a solution architect worked with the Swarm team on a design for the AEM solution.
Opencast’s business analysts worked with the PO and ADM to engage and advise Swarm on service delivery timelines and milestones to align expectations and ensure clarity on delivery. Deliverables from a BA perspective included a customer onboarding journey, front-end visualisation and battery management. The team also conducted accessibility reporting on the web app.
Security considerations were paramount. Opencast’s Head of Security Services helped Swarm to enhance its cybersecurity controls and receive certification for compliance, including new government legislation brought in April 2024, enabling Swarm to achieve IoT Baseline Level 1 assurance provided by certifying body IASME.
Work included adopting secure coding practices, conducting regular security audits and third-party penetration tests and implementing robust data protection protocols. By focusing on risk assessment and management, Swarm was able to identify and mitigate potential vulnerabilities early in its development cycle.
Cross-cutting work including in product, service design and delivery is enabling Swarm to scale the new service, helping it to deliver its first few hundred orders. And a scalability plan was put in place to meet future projections.
In spring 2024, the new product was rolled out for user and technical testing to 15 homes. It assessed users’ individual tariffs, historical energy demand and solar generation, and automated their buying and selling of energy from the grid. The AEM would automatically make decisions on behalf of customers to buy and sell energy and show them the financial savings they made each day.
In July 2024, Swarm’s Hub:One product went live, offering a UK first – an automated home energy management system that worked across renewable devices from different providers. Later that month, Swarm secured its first 500 units to a key wholesaler in the marketplace. It aimed for at least 3,000 devices to be installed by the end of 2024.
The partnership also refined and accelerated Swarm’s go-to-market strategy, including identifying market opportunities that would reinforce the unique selling point of Swarm’s device.
Early impact from a 15-home roll-out showed a big drop in the time for people to see a return on their investment in renewable energy.
With a solar setup, one bill was reduced to £12, and with the AEM device installed the total monthly bill was zero through a combination of savings and selling excess energy back to the grid at peak times. Customer homes using solar were also becoming net positive.
The deployment of the Hub:One technology, with its orchestration of multiple systems in under one second, paves the way for future functionality including in the virtual power plant. This technology has the potential for other applications in future.
Swarm’s Hub:One product passed the government-set IoT product security and telecommunications infrastructure (PSTI) requirements.
Swarm’s next-gen technology, developed with the support of Opencast, has a significant impact on home energy consumption, transforming how consumers manage, store and redistribute energy, with a single device making intelligent decisions on their behalf. And it accelerates their payback on renewables investment. The algorithm at its centre understands users’ needs, solar forecasts and energy tariffs.
Consumers no longer need to make home energy decisions: Swarm does it all for them.
up to
savings on energy bills
units shipped early to key wholesaler
up to
new North East jobs
Cloud first: using IoT technologies and serverless AWS monitoring of live services
Kotlin APIs: to surface data and informationacross the platform
Python, SQL, OR Tools and Databricks: for data science and automation algorithms
Typescript: used to create software for the Hub:One device to talk to the inverters
Custom hardware and Modbus: to design and prototype hardware
React: for front-end development
Third-party APIs: to integrate withelectric vehicles and chargers
Mobile apps: designing for iOS and Android
Jira, Confluence, Figma, Miro and Excalidraw: for collaboration.
Fill out the short form and we’ll email you a link to download a PDF of the case study.
Loading...