Exploring Quantum Algorithms for Optimal Sensor Placement in Production Environments
summary
Large-scale sensor-placement is encoded as a QUBO and solved with quantum annealing with D‑Wave and gate-model QAOA via IBM and Pennylane, benchmarked against Gurobi and CPLEX. Real factory instances require ~4,100 logical qubits; decomposition into eight subproblems of ~600–700 variables enables hardware runs. Hyperparameter tuning finds penalty weights below 1 and longer anneal times improve results; decomposition yields feasible near-optimal layouts. Classical solvers still outperform on constrained instances, but quantum methods show industrially relevant cost savings potential (€5,000/sensor → >€1M/yr).
AI-written summary · about
chapters (8)
- 0:00Quantum sensor placement project
- 1:28Optimal sensor placement challenges
- 4:12Quantum approach to sensor placement
- 7:52Hyperparameter optimization for annealing
- 10:39Decomposition techniques for scaling
- 12:41Quantum vs classical constraints
- 14:52Custom quantum annealing approaches
- 15:34Conclusion and future work
related talks
41mQuantum Computing in Practice: Where We Are, What Works, and What’s Next
21mSoftware for Quantum Computing- Easing Solutions for Complex
19mQuantum Technology in Flexible Job Shop Scheduling?
47mUnboxing Quantum Computing
22mSchrödinger’s Cat in Industry: Real Life Quantum Software Engineering
19mExploring Quantum Advantage for the German industry
14mQuantum Computing- Exploring Automotive Applications
21mThe strategic relevance of quantum computing for logistics companies