Public organisations increasingly procure digital solutions, including cloud services, software platforms, data systems, AI tools, and digital infrastructure. Such solutions can improve service quality, efficiency, transparency, and innovation in the public sector. However, they also introduce cybersecurity risks related to data protection, system vulnerabilities, supplier access, operational continuity, and dependence on external technology providers.

Cybersecurity requirements are therefore becoming an important part of tender design and contract management in public procurement. Public buyers must ensure that digital solutions are secure and compliant with relevant regulations and standards. At the same time, cybersecurity requirements that are too strict, vague, or poorly designed may reduce competition, discourage innovative suppliers, increase costs, or exclude smaller firms with valuable solutions. This creates an important procurement challenge: how can public buyers control cybersecurity risks while still encouraging innovation in digital public services?

A master’s thesis on this topic may explore how cybersecurity considerations are incorporated into the public procurement of digital solutions, and how these requirements influence innovation, supplier participation, and procurement outcomes.

NB: This topic is supported by Anskaffelsesakademiet – the Norwegian Academy of Public Procurement. Students who choose this topic may receive up to NOK 10,000 to support their research. In addition, students who receive an A or B on their final thesis assessment may be selected to participate in the national competition for the best master’s thesis in public procurement in Norway. The competition includes students from universities and university colleges across the country, and the winning thesis receives NOK 30,000. As inspiration, the award has been offered four times so far, and students from Molde University College have won it twice.

Keeping materials and products in use sits at the heart of the circular economy, but a firm can only manage what it can measure. Data has become the center of gravity for digital circular solutions: a review of 266 studies on digital enablers found that 89% involved collecting, storing, analyzing, or sharing data [1]. Sensors, shared ledgers, and analytics now make it possible to follow a product through its life and feed that record back into decisions about reuse, repair, remanufacturing, and resale [2]. The digital product passport, which the EU is introducing across several product groups, is one route to that data [3].

The promise is real-time visibility; the practice lags behind. Much product data is collected and stored, yet little of it changes what a firm does at the point of decision. A review of the electronics sector identified 77 distinct information barriers across the product life cycle, including missing material composition information, dismantling instructions for recyclers, and product condition data [4]. It is often unclear which fields matter, who acts on them, and what blocks their use.

In this master's thesis, the student will investigate how product-level data supports circular operating decisions in one value chain and what stops firms from acting on it.

• Which data fields in a digital product passport actually change a firm's reuse, repair, or remanufacturing decision, and which go unused?

• What organizational and technical barriers stop firms from acting on real-time product data?

• How does access to predictive analytics change material-flow and inventory decisions compared with rule-based reordering?

Potential cases: value chains in electronics, automotive, or construction, where digital product passports are being introduced.

Suggested approach: Mixed: Interviews or a case study to surface which product data firms act on and why, then a survey analyzed with PLS-SEM.

References

[1] Digital Enablers of the Circular Economy: A Systematic Review (2026). Journal of Manufacturing and Materials Processing, 10(4), 112. https://doi.org/10.3390/jmmp10040112

[2] Wilson, M., Paschen, J., & Pitt, L. (2022). The circular economy meets artificial intelligence (AI). Management of Environmental Quality, 33(1), 9-25. https://doi.org/10.1108/MEQ-10-2020-0222

[3] Digital product passports as enablers of the digital circular economy (2024). Telecommunications Systems. https://doi.org/10.1007/s11235-024-01104-x

[4] Information Barriers to Circularity for Electronic Products and the Potential of Digital Product Passports (2025). Sustainability, 17(12), 5554. https://doi.org/10.3390/su17125554

Research on digital technologies for the circular economy tends to study one tool at a time, blockchain here, IoT there, AI somewhere else. Real circular solutions rarely work that way. A traceable, closed-loop supply chain usually requires sensors to capture data, a shared ledger to ensure trustworthiness, and analytics to act on the data, with each component covering the others' weak spots [1].

The evidence on combinations is starting to appear. An empirical study across platform-based sectors found that blockchain integration, AI, and IoT each increase supply-chain transparency, which in turn improves circular supply-chain performance, and that the strongest lever varies by sector, with blockchain mattering most in urban mobility and IoT having the largest direct effect in consumer electronics [2]. What is still thin is the cost side: how much the combination adds over a single technology, and whether the extra complexity pays off. Digital twins add an additional layer, enabling firms to model circular operations and industrial symbiosis before committing resources [3,4].

In this master's thesis, the student(s) will investigate how two or more digital technologies, used together, affect performance in a multi-tier circular supply chain.

• How do pairing technologies (for example, IoT tracking with blockchain traceability) change transparency and product-recovery rates compared with using either one alone?

• What are the cost-benefit trade-offs of deploying combined digital technologies for circularity in a chosen sector?

• How do digital twins support adaptive circular manufacturing or industrial symbiosis in practice?

Potential cases: a manufacturer or logistics provider running more than one digital technology in the same chain.

Suggested approach: Quantitative: PLS-SEM plus cost-benefit modeling for the trade-off question.

References

[1] Digital transformation in supply chains: improving resilience and sustainability through AI, Blockchain, and IoT (2025). Frontiers in Sustainability. https://doi.org/10.3389/frsus.2025.1584580

[2] Xing et al. (2025). Enhancing Circular Supply Chain Performance in Platform-Based Economies. Corporate Social Responsibility and Environmental Management. https://doi.org/10.1002/csr.70274

[3] Digital twins for environmentally sustainable and circular manufacturing sector (2024). Production & Manufacturing Research, 12(1). https://doi.org/10.1080/21693277.2024.2428249

[4] A Review of Digital Twin Integration in Circular Manufacturing for Sustainable Industry Transition (2025). Sustainability, 17(16), 7316. https://doi.org/10.3390/su17167316

More circular activity is driven by digital platforms. Instead of selling a product once, firms use platforms to share, lease, or offer it as a service, manage it throughout their life, and open second-hand and resource-exchange markets [1]. The platform serves as the coordinator for a broader ecosystem of users, suppliers, and service partners. A systematic review of 134 studies on circular-economy platforms maps this field but finds the evidence still fragmented and largely descriptive [2].

Coordination is the hard part. A platform only works if participants trust it and continue to take part, which depends on how it is designed and governed. The financial case is often unsettled too: circular business models can carry high upfront investment and unclear timelines for returns [3].

In this master's thesis, the student(s) will investigate how a digital platform coordinates a circular business model and what keeps participants trusting and active.

• How do digital platforms make sharing, leasing, or product-as-a-service models work in a specific market?

• What governance mechanisms build trust and sustain participation in a platform-based circular network?

• How do platform design choices affect consumer willingness to choose circular options over ownership?

Potential cases: a sharing, leasing, or product-as-a-service platform in a chosen market.

Suggested approach: Mixed: Qualitative case study and interviews for the governance and trust questions; a survey analyzed with PLS-SEM.

References

[1] OECD (2019). Business Models for the Circular Economy: Opportunities and Challenges for Policy. OECD Publishing, Paris. https://doi.org/10.1787/g2g9dd62-en

[2] Blackburn et al. (2026). Circular Economy Platforms: A Systematic Review. Business Strategy and the Environment. https://doi.org/10.1002/bse.70307

[3] Balancing Economic Viability and Environmental Impact in Circular Business Transitions (2026). Circular Economy and Sustainability. https://doi.org/10.1007/s43615-026-00968-2

As governments increasingly integrate digital tools into procurement processes, DPPs—structured datasets that provide standardized, lifecycle-based information about products—are gaining policy traction, particularly within the EU’s circular economy framework and the proposed Ecodesign for Sustainable Products Regulation (ESPR). Their relevance lies in enabling contracting authorities to access more reliable and comparable information on environmental performance, material composition, and supply chain characteristics, which can strengthen green public procurement practices and address persistent information asymmetries between buyers and suppliers. At the same time, digital transparency tools such as DPPs may incentivize firms to innovate by aligning product design and production processes with new data disclosure and sustainability requirements. However, their implementation raises challenges related to data standardization, interoperability, governance structures, and varying levels of supplier readiness, particularly among SMEs.

Aim: To examine how digital product passports can be integrated into public procurement processes and to assess their implications for innovation and sustainability outcomes.

NB: This topic is supported by Anskaffelsesakademiet – the Norwegian Academy of Public Procurement. Students who choose this topic may receive up to NOK 10,000 to support their research. In addition, students who receive an A or B on their final thesis assessment may be selected to participate in the national competition for the best master’s thesis in public procurement in Norway. The competition includes students from universities and university colleges across the country, and the winning thesis receives NOK 30,000. As inspiration, the award has been offered four times so far, and students from Molde University College have won it twice.

This thesis topic focuses on how Wenaas can optimize inventory allocation between its main warehouse in Norway and its international warehouses. The students may analyze sales history, demand patterns, product characteristics, lead times, service requirements, and inventory turnover to identify which products should be stocked locally in international markets and which products should be served from the central warehouse in Norway. The study may also consider trade-offs between availability, cost, responsiveness, working capital, and logistics complexity.

This thesis topic focuses on identifying which areas of Wenaas’ purchasing processes can be automated through the use of digital technologies. The students may map current purchasing activities, identify repetitive or time-consuming tasks, and assess where automation could improve efficiency, accuracy, control, or decision-making. The study may include a cost-benefit assessment of potential automation initiatives and, if relevant, provide a deeper analysis of one or two high-priority initiatives with clear implementation potential.

This thesis topic focuses on how organizations can strengthen security and preparedness through improved procurement processes, supplier management, and supply chain management. Students may examine how risks related to critical suppliers, supply security, dependencies, vulnerabilities, and emergency preparedness can be identified and managed in public or private procurement. The thesis may also explore how requirements related to security, resilience, and preparedness can be integrated into tender documents, contracts, supplier follow-up, and strategic purchasing decisions.

Manufacturers holding a wide range of components face a recurring question: which items should be kept in the main warehouse and which in a secondary one closer to where they are needed. The choice shapes how fast orders are filled, how much stock is tied up, and what the whole network costs. The idea that coordinated decisions across stocking levels beat node-by-node decisions goes back to foundational multi-echelon inventory theory [1].

Adding a secondary warehouse changes lead time, service level, and total cost simultaneously, and these pull in different directions, where to place stock across a network is itself a well-studied optimization problem [2]. In practice, this decision is further complicated by uncertain demand, variable lead times, and differences in the criticality of individual components to operations. The wider literature on multi-echelon inventory under uncertainty is large; one review classifies 394 studies, yet a clean, practical rule for splitting a real component range between two warehouses remains case-specific [3].

In this master's thesis, the student(s) will develop and test a method for allocating components between a main and a secondary warehouse, and measure how the secondary warehouse affects location, lead time, service level, and total cost. The thesis will also examine suitable replenishment policies for the two-warehouse setting, including how ordering rules and stock control parameters should be adapted when components are stored at different levels of the network.

· Which component characteristics, such as demand volume, demand variability, value, criticality, and replenishment lead time, should drive the choice between main and secondary warehouse storage?

· How does adding a secondary warehouse change lead time, service level, inventory levels, and total cost across the network?

· Which allocation rule gives the best balance between service level and total cost for the studied product range?

· Which replenishment policies are most suitable for components stored in the main and secondary warehouses, and how should their parameters be adapted to the two-warehouse setting?

Company: Brunvoll.

Suggested approach: EOQ, reorder point systems, safety stock, ABC/XYZ classification, two-echelon inventory allocation.

References

[1] Clark, A. J., & Scarf, H. (1960). Optimal Policies for a Multi-Echelon Inventory Problem. Management Science, 6(4), 475-490. https://doi.org/10.1287/mnsc.6.4.475

[2] Graves, S. C., & Willems, S. P. (2000). Optimizing Strategic Safety Stock Placement in Supply Chains. Manufacturing & Service Operations Management, 2(1), 68-83. https://doi.org/10.1287/msom.2.1.68.23267

[3] de Kok, T., Grob, C., Laumanns, M., Minner, S., Rambau, J., & Schade, K. (2018). A typology and literature review on stochastic multi-echelon inventory models. European Journal of Operational Research, 269(3), 955-983. https://doi.org/10.1016/j.ejor.2018.02.047

Safety stock is the buffer that protects service when operations do not go as planned. In classical inventory control, safety stock is often dimensioned mainly based on demand variability, while replenishment lead times and supplier performance are treated as stable. In practice, however, manufacturers may face uncertainty on both sides of the inventory system: customer demand may fluctuate, and suppliers may deliver late, partially, or with variable lead times.

From an operations research perspective, this problem belongs to stochastic inventory control, with links to service-level constrained inventory optimization, lead-time demand modelling, demand forecasting, and safety-stock positioning in multi-echelon systems. Models that account for stochastic lead times show that supplier reliability can substantially affect the buffer stock required to maintain a target service level [1]. When demand and supply uncertainty are present, a demand-only safety-stock rule may underestimate the true risk of stockouts.

A further important modelling question is how demand itself should be represented. In a practical inventory component, demand may be regular, seasonal, intermittent, or highly variable, and different demand patterns may require different forecasting tools and safety-stock rules. Statistical forecasting methods, demand classification approaches, or data-driven prediction tools can therefore be used to estimate future demand and demand uncertainty before safety stock is dimensioned.

The problem is also important from a logistics perspective because safety stock affects not only service level, but also inventory investment, replenishment frequency, warehouse capacity, and total logistics cost. Where to hold the buffer and how much to hold can therefore be interpreted as a safety-stock dimensioning and positioning problem within a supply network [2]. The broader literature on stochastic multi-echelon inventory models provides several modelling approaches but translating these into practical rules that combine demand prediction, supplier reliability, and safety-stock dimensioning remains a relevant applied challenge [3].

In this master's thesis, the student(s) will build and test a safety-stock model that accounts for both demand variability and supplier-reliability uncertainty, supported by appropriate demand modelling or prediction tools, and compare its performance with current practice.

· How should demand be modelled or predicted for the component range studied, and how do different demand patterns affect safety-stock requirements?

· How should safety stock be dimensioned when both demand variability and supplier lead-time uncertainty are present, and how does this differ from demand-only models?

· How can supplier reliability, such as lead-time variability, delivery delays, or delivery performance, be incorporated into safety-stock calculations?

· How sensitive is the required safety stock to changes in supplier reliability, demand variability, and forecasting accuracy?

· What service-level, inventory-investment, and total-cost outcomes result from the proposed model compared with current practice?

Company: Brunvoll.

Suggested approach: Demand forecasting, ABC/XYZ classification, reorder point systems, safety stock under stochastic lead times, supplier reliability analysis.

References

[1] Simchi-Levi, D., & Zhao, Y. (2005). Safety Stock Positioning in Supply Chains with Stochastic Lead Times. Manufacturing & Service Operations Management, 7(4), 295-318. https://doi.org/10.1287/msom.1050.0087

[2] Graves, S. C., & Willems, S. P. (2000). Optimizing Strategic Safety Stock Placement in Supply Chains. Manufacturing & Service Operations Management, 2(1), 68-83. https://doi.org/10.1287/msom.2.1.68.23267

[3] de Kok, T., Grob, C., Laumanns, M., Minner, S., Rambau, J., & Schade, K. (2018). A typology and literature review on stochastic multi-echelon inventory models. European Journal of Operational Research, 269(3), 955-983. https://doi.org/10.1016/j.ejor.2018.02.047

Forecasting demand is challenging for any new product, and even more so when there is little or no sales history to rely on. Traditional time-series forecasting methods are usually built on repeated observations of past demand, but new products often face a cold-start situation where such data are missing, sparse, or highly unstable. Reviews of new-product forecasting describe this problem, absent historical data and volatile early demand, and point to data-driven and machine-learning methods as the way forward [1]. Recent work has further shown how deep learning and other data-driven methods can support forecasting for newly launched products with short life cycles [2], while pre-launch signals and customer-response data can also improve forecasts before substantial sales history is available [3].

From an operations research and logistics perspective, this problem belongs to demand forecasting and inventory planning under uncertainty, with links to cold-start forecasting, machine learning, probabilistic forecasting, and procurement decision support. Instead of relying solely on early sales observations, modern forecasting approaches can leverage additional information, including product attributes, customer segments, market signals, comparable products, expert judgment, web traffic, pre-orders, and campaign data. Machine-learning methods are particularly relevant because they can learn demand patterns from related products or external features and transfer this knowledge to a new product with limited history.

The demand pattern considered in this thesis is also of practical interest. An early-stage company may face two different types of demand at the same time: a short, sharp demand window linked to a niche use case, and a steadier long-term demand associated with ordinary use. These two patterns may require different forecasting logic. The short-window demand may depend on timing, awareness, events, or specific customer needs, while ordinary-use demand may be better represented by gradual adoption, repeat demand, or market growth. A single aggregate forecast may therefore hide important differences between temporary and stable demand components.

Recent developments in machine learning and time-series forecasting offer several possible directions for this type of problem. These include feature-based forecasting, similar-product or analogue-based forecasting, tree-based machine-learning models, probabilistic forecasting, and newer transfer-learning or foundation-model approaches for time series. The purpose is not only to produce a point forecast, but also to quantify forecast uncertainty and translate it into procurement and inventory decisions. This is important because overestimating demand can tie up scarce cash in excess stock, while underestimating demand can lead to missed sales during a narrow demand window.

In this master's thesis, the student will test forecasting approaches suited to a new product with little sales history and two distinct demand patterns. The study should compare traditional forecasting logic with selected machine-learning or data-driven approaches and evaluate how forecast uncertainty should inform procurement and inventory decisions.

· Which forecasting methods are most suitable for a new product with little or no sales history?

· How can machine-learning methods use product attributes, comparable products, market signals, or early demand observations to improve cold-start demand forecasts?

· How can the company distinguish between short-window niche demand and steady ordinary-use demand?

· How should forecast uncertainty be quantified and incorporated into procurement and inventory decisions?

· What are the cost and service implications of using the proposed forecasting approach compared with simpler forecasting or judgment-based methods?

Company: Carry On.

Suggested approach: Cold-start demand forecasting, demand segmentation, machine-learning models, probabilistic forecasting, inventory decision support.

References

[1] S., A., & R., N. (2025). Demand Forecasting New Fashion Products: A Review Paper. Journal of Forecasting, 44(2), 270-280. https://doi.org/10.1002/for.3192

[2] Elalem, Y. K., Maier, S., & Seifert, R. W. (2023). A machine learning-based framework for forecasting sales of new products with short life cycles using deep neural networks. International Journal of Forecasting, 39(4), 1874–1894.

[3] Harz, N., Hohenberg, S., & Homburg, C. (2022). Virtual Reality in New Product Development: Insights from Prelaunch Sales Forecasting for Durables. Journal of Marketing, 86(3), 157–179.

Moving from a working prototype to commercial volumes is a critical transition phase for many product start-ups. At the prototype stage, the supply chain is often informal, low-volume, and built around flexibility and experimentation. However, commercialization requires a more structured supply chain network that can support higher volumes, stable quality, reliable deliveries, and controlled costs. Designing this transition too early may lock the company into fixed capacity, unsuitable suppliers, or costly infrastructure; designing it too late may prevent the company from meeting market demand when growth begins.

From a logistics and operations research perspective, this problem can be understood as a supply chain network design and scale-up planning problem under demand and capacity uncertainty. The company must decide how the supply chain should be structured, which activities should be performed internally or outsourced, which suppliers to select, and how sourcing and capacity decisions should evolve as demand increases. These decisions are interdependent: supplier choices affect cost, lead time, quality, flexibility, and scalability, while network design choices determine how easily the company can transition from prototype production to commercial operations.

The practical challenge is to design a supply chain that remains cost-effective at low early volumes while remaining scalable as demand grows. This requires balancing short-term efficiency with long-term flexibility. A low-cost structure may be attractive during early commercialization, but it may become a bottleneck if suppliers, production capacity, or distribution arrangements cannot expand. Conversely, investing too early in large-scale capacity or complex logistics may create unnecessary fixed costs before demand is proven. Evidence on circular business transitions indicates that start-ups must scale and establish themselves while still building their capabilities [1], making staged and adaptable supply chain designs particularly relevant.

Supplier and sourcing decisions are therefore central to the problem. Early choices regarding local versus global sourcing, single versus multiple suppliers, outsourcing, minimum order quantities, supplier reliability, and contract flexibility can strongly influence both cost and scalability [2]. A suitable supply chain design should not only identify a feasible structure for the current prototype-to-market phase but also evaluate how this structure can evolve through different demand-growth scenarios.

In this master's thesis, the student will propose and test a supply chain design that supports the transition from prototype to commercial scale. The study should evaluate alternative supply chain structures with respect to cost efficiency, scalability, supplier and sourcing choices, and risk exposure.

· What supply chain network structure best supports the transition from prototype production to commercial-scale operations?

· How can the supply chain remain cost-effective at low early volumes while still allowing scalable growth?

· Which supplier and sourcing strategies best support cost, flexibility, quality, and future volume expansion?

· How sensitive is the proposed supply chain design to different demand-growth and capacity scenarios?

· What are the main risks in the scale-up path, and how can the supply chain design absorb or mitigate them? 

Company: Carry On.

Suggested approach: Supply chain network design, supplier selection, capacity planning, demand-scenario simulation, mixed integer programming.

References

[1] Balancing Economic Viability and Environmental Impact in Circular Business Transitions (2026). Circular Economy and Sustainability. https://doi.org/10.1007/s43615-026-00968-2

[2] A Review of Sustainable Supplier Selection with Decision-Making Methods from 2018 to 2022 (2024). Sustainability, 16(1), 125. https://doi.org/10.3390/su16010125