CETO Innovation
Predictive maintenance for district heating networks
/images/ceto-innovation/hero.webp
Receiving the third-place award at DTU Green Challenge, holding the framed prize plaque
- Client
- DTU Skylab Incubator
- Context
- X-Tech Entrepreneurship, DTU
- Year
- 2025
- Role
- Industry & Partnerships
- Team size
- 7
- Tools
- Market sizing · Unit economics · Customer interviews · budget and forecasts
- Contribution
- Team of 7, industry outreach, market model, customer validation and forecasting
A predictive maintenance SaaS for district heating networks, taken from a course concept into the DTU Skylab incubator, with roughly 3 M DKK in annual saving for a medium-sized utility, a 320.85 M DKK addressable market and third place at DTU Green Challenge.
The decision behind the venture
Utilities do not buy on prediction accuracy alone, and talking to Danish district heating operators showed that replacement is decided by budget cycles, dig permits and political risk rather than by a probability number. This means that a model which is 85% accurate instead of 65% only sells if it produces a defensible maintenance plan that a utility can put in front of a board, and that evidence pushed the product from an accuracy story to a capital-planning story.
What I owned
I held the industry and partnership side of a seven-person venture, which meant reaching out to Danish district heating utilities, running the conversations that showed how maintenance decisions are actually made today, and turning that into the market and customer case, covering TAM, SAM and SOM from the CVR register and Nordic network data, the unit economics, and the benchmark against Direct C, Solinas, Integrity and Preventio. It is important to note that the machine learning was Felicia and David's, the software Sólon's and the infrastructure Anastasia's, while Nayeli ran the project and Liza was CEO. Here my job was making sure that what they built matched what utilities would actually buy.
The problem
District heating utilities replace pipe on schedule rather than on evidence, and this wastes capital while still missing the failures that matter.
The solution
A platform combining historical, operational and environmental data in order to give real-time failure risk predictions, which moves the prediction accuracy from 65% to 85%.
The customer case
Roughly 3 M DKK in annual savings for a medium-sized utility, achieved through optimised maintenance scheduling rather than through replacing more pipe.
The market
TAM 1,350 bn DKK, SAM 320.85 M DKK and SOM 173.1 M DKK, built from the Danish CVR register and Nordic network data, with unit economics of 8 k DKK in customer acquisition cost against 918 k DKK in profit on a single sale. Furthermore the offering was benchmarked against Direct C, Solinas, Integrity and Preventio.
/images/ceto-innovation/dashboard.webp
Heat Network Monitoring dashboard: a street map with pipe segments colour-coded by condition
/images/ceto-innovation/market-need.webp
Stat graphic: of 400 district heating companies in Denmark, 50% want to improve current operations
/images/ceto-innovation/maintenance-schedule.webp
CETO Innovation dashboard mockup showing a scheduled maintenance list with priority levels and system suggestions
/images/ceto-innovation/pipes-list.webp
CETO Innovation dashboard mockup: a network pipes table listing pipe ID, condition, type, material, diameter, length and risk score