Conference Agenda

The COVID-19 pandemic has significantly impacted the organization and attendance of live events for several years, particularly in confined spaces such as arenas. To address these challenges, we developed and analyzed models for optimal arena seating configurations that adhere to social distancing protocols. Our primary focus is on maximizing seating capacity while ensuring safety through appropriate spacing between attendees from different households.

Our approach involves formulating a 2D knapsack problem-inspired model, where seats are allocated to groups of varying sizes, considering both horizontal and vertical distancing constraints. This model enables the generation of seating plans that optimize the number of attendees and comply with health regulations. We explored two distinct strategies: operational/tactical and strategic, each tailored to different event types and priorities, such as maximizing overall attendance or accommodating VIP and sponsor groups.

Our collaboration with a German football club demonstrates the effectiveness of our models. By applying our approach under various scenarios and constraints, we demonstrated their adaptability and potential for substantial gains in seating capacity and event revenue compared to traditional methods. Our framework offers a robust decision-making tool that can be easily extended to other domains requiring similar social distancing protocols.

For a system whose state declines when ’active’ and increases otherwise, we determine optimal work-rest cycles. The optimal switching policy maximizes the average cycle benefit, where the benefit at any given time is proportional to the system state during an active period and zero otherwise. Any given lower bound for the length of a rest period determines a unique limit cycle, where the state converges, irrespective of its initial value. The average benefit becomes maximal when the length of the resting period converges to zero and the system reaches a steady state of ’flow’ with a nontrivial work-rest split. For that case, we provide a relatively robust parameter estimate, valid in the absence of any knowledge about the system’s time constants. The results apply to regenerative systems determining the optimal split between uptime and downtime for maximum user benefit.