Avinor’s AI Takeoff: Building Internal AI Capabilities

Avinor's mission to connect Norway and the world through sustainable aviation requires a commitment to continuous operational and financial improvement. That includes investment in emerging technologies. Recognizing the potential of artificial intelligence (AI) to enhance efficiency, Avinor partnered with BCG to explore how AI could create value for its airport operations.

Setting the Course for Avinor's AI Journey

Rapid and tangible results were a top priority. At the same time, Avinor’s leadership team required a long-term AI strategy that fully leveraged the technology’s potential. Avinor also wanted its AI journey to be self-funding so that each phase of the project would deliver value to fund continued investment, generate internal momentum, and secure executive buy-in. Finally, Avinor insisted that its reliance on third-party expertise would be temporary, and that the initiative would result in a robust internal AI capability. To that end, the implementation process was designed to be a phased transition that would first build the capabilities in Avinor and then enable the Avinor team to gradually assume full ownership of deployment.

After identifying and prioritizing a set of AI use cases, BCG and Avinor worked together to develop several proofs of concept, with the Avinor team taking on an increasing amount of responsibility. In addition to demonstrating AI’s ability to deliver tangible value, this collective effort established Avinor’s internal AI capabilities and provided a basis for expanding AI applications across the entire ecosystem of airport operations.

To digitize and future-proof operations, Avinor envisioned AI as a business-driven, strategic enabler that enhances rather than replaces human skills. In selecting a partner, Avinor was drawn to BCG’s blend of strategic and technical expertise, as well as the ability to deliver business value and lead a truly collaborative implementation.

Laying the Groundwork for Lasting AI Capabilities

The 12-month engagement began with BCG helping Avinor establish a foundation for its internal AI capabilities. This involved defining the roles and responsibilities, governance mechanisms, organizational structures, data maturity frameworks, and partnerships required to enable knowledge transfer from BCG to Avinor. BCG also supported initial recruitment efforts to build Avinor’s internal AI team. Throughout this process, fostering strong collaboration between the newly established AI team, the business units, and executive leadership was critical to success.

Subsequently, a rapid assessment of the operational landscape identified more than 60 potential AI use cases across the airport ecosystem. The team then assessed each use case for value and feasibility across 12 dimensions - such as bottom-line impact, risk and cost avoidance, innovation, and customer experience.

Proving the Value of AI Through Early Success

The assessment resulted in a list of prioritized use cases. The first, Automatic Invoice Control (AIC), aimed to streamline invoicing management processes. AI capabilities deployed to deliver the proof of concept (PoC) for this use case included optical character recognition to extract contract price tables, fuzzy matchings and embeddings to detect outliers and terms, and GenAI functionality to facilitate invoice claims processing.

The assessment of the PoC identified significant value opportunities, indicating that a portion of the spend may not be fully optimized at present. Additional benefits included streamlined workflows and more accurate data that could be leveraged in contract negotiations. This initial step, moreover, built confidence among leadership and employees, and established the credibility of AI’s ability to impact the bottom line.

Embedding AI in the Heart of Airport Operations

The second PoC developed an AI-driven model to optimize the operationally complex task of allocating stands and gates around arriving (and departing) aircraft. Over the course of ten weeks, BCG consultants - working shoulder to shoulder with the Avinor team - implemented a model that processed the most important of the 20,000 total soft and hard allocation rules using Monte Carlo simulation, machine learning, and optimization programming. The outcome of the model showed a potential value uplift and that it is possible to add an extra flight during peak hours without losing operational robustness.

Another PoC tackled a quintessentially Norwegian challenge: harsh winter weather. Here, the Avinor team - now working independently after upskilling from BCG’s team - developed an AI model that gave snow and ice removal teams better data for making decisions on when and how much anti-icing chemicals to apply to runways. The initiative showed that improved decision support can reduce chemical usage by 10% to 20%.

Scaling AI Impact Across the Entire Avinor Network 

Each of the PoCs delivered tangible benefits. More importantly, the collaborative effort built a foundation of internal expertise, improved operating models, and enhanced governance structures and processes. This foundation positioned Avinor to continually expand AI capabilities across the entire ecosystem of its operations.

While numerous opportunities exist to apply AI across Avinor’s operations, it will be essential to avoid isolated, one-off technology solutions. Instead, success will depend on embedding AI into Avinor’s culture, ways of working, and decision-making - ensuring that innovation becomes a sustained, organization-wide capability. In BCG, we see that successful, company-wide AI transformation typically follows a 10-20-70 logic where success is 10% dependent on AI algorithms, 20% on broader technological topics, and 70% on succeeding with people and processes. Avinor subsequently shifted its focus accordingly, and has invested in building dedicated AI adoption capabilities that focus on upskilling and bringing AI into day-to-day work.

Building Confidence and Ownership Through Hands-On Experience

While BCG executed the initial PoC, the two teams collaborated on the second. After doing the preparatory work on the third PoC, BCG handed execution off to the Avinor team, which implemented the fourth PoC entirely on its own. This structured transition from discrete “PoC pilots” to embedded AI expertise demonstrated the viability of Avinor’s strategy and enhanced credibility across the organization.

The Next Destination for Avinor

At the end of the year-long engagement, Avinor achieved its key goal of building a robust internal AI capability. Upskilling, along with a rigorous recruitment effort that attracted high-quality hires, embedded AI skills within the organization. The PoC implementation process also enabled the integration of essential AI tools, governance frameworks, and data models.

These developments have helped to position Avinor as a forward-looking, technologically ambitious operator that has embraced AI as a pillar of its future competitiveness. By demonstrating AI’s ability to deliver both cost savings and revenue growth, moreover, the initiative has generated broader awareness and enthusiasm across business units.

BCG provided global expertise and frameworks, delivered PoCs, anchored the work within business units and leadership, and transferred capabilities to build Avinor’s long-term independence. Through a tailored program deployed over three phases, BCG established the foundational capabilities needed to kick off Avinor’s AI journey, delivered rapid value through supporting PoC deployment, and created a future-fit internal AI organization equipped to tackle the full range of opportunities presented by airport operations.

For Avinor, the collaboration went beyond a successful AI pilot project to establish a foundation for a sustainable AI-driven transformation.