Operations research (OR) is a scientific approach to analyzing complex operational systems and making decisions to improve performance. OR techniques like linear programming, queuing theory, and simulation have been used since World War II to optimize everything from supply chains to airline scheduling. However, with the rise of “big data” and advanced analytics, some wonder if old-school operations research is still relevant.
What is operations research?
Operations research, also known as operational research or optimization science, is an interdisciplinary field that applies advanced analytical methods like mathematical modeling to help make better decisions. The goal is to provide rational bases for decision making by seeking to understand and structure complex situations and identify optimal solutions to operational problems.
Some of the key tools and methodologies of operations research include:
- Linear programming – Optimizing the outcome of a mathematical model whose requirements are represented by linear relationships
- Integer programming – A variant of linear programming where variables are restricted to integer values
- Nonlinear programming – Optimizing an outcome with nonlinear objective and/or constraint functions
- Markov decision processes – Modeling sequential decision making in situations where outcomes are partly random and partly under the control of a decision maker
- Queuing theory – Analysis of waiting lines (queues)
- Game theory – Modeling strategic situations where the optimal outcome depends on the choices of multiple decision makers who may have conflicting interests
- Stochastic modeling – Quantifying uncertainty and risk
- Simulation – Mimicking the operation of real-world systems and processes
These and other OR techniques allow analysts to model complex situations and find optimal or near-optimal solutions. Operations research is used in a wide range of industries and disciplines, including manufacturing, transportation, energy, telecommunications, finance, logistics, healthcare, retail, and government services.
When did operations research originate?
Operations research originated in Britain during World War II. In 1937, English scientist Patrick Blackett used statistical analysis to improve military operations, coining the term “operational research.” During the war, groups of scientists solved strategic problems related to the effective operation of weapon systems and military units using analytical methods.
In 1948, the Operations Research Society of America was founded to promote the development and practice of OR. In 1952, Philip M. Morse and George E. Kimball published Methods of Operations Research, the first textbook on OR, popularizing it as an academic discipline. The construction of early computers expanded the implementation of OR methodologies.
Since the war, operations research has grown into an established technical discipline with applications in many fields. It is widely used in large organizations for decision support, logistics, and resource allocation.
What are some applications of operations research?
Here are some examples of where operations research is commonly applied:
- Transportation – Optimizing routes and schedules for airlines, trains, trucks, ships, etc. Planning transportation networks.
- Manufacturing – Production planning, inventory control, supply chain optimization, assembly line scheduling, quality control.
- Energy – Optimizing power generation systems. Planning electricity transmission networks. Designing renewable energy systems.
- Telecommunications – Designing cell tower networks. Call routing optimization.
- Retail – Inventory management, supply chain optimization, warehouse operations, workforce scheduling.
- Finance – Portfolio optimization, risk management, algorithmic trading systems.
- Healthcare – Hospital operations, appointment scheduling, facility layout planning, public health optimization.
- Government – Resource allocation, disaster preparedness and relief, nuclear waste disposal, border control optimization.
Operations research techniques allow organizations to maximize output and efficiency in complex operational environments and processes. The applications are virtually endless across the public and private sectors.
What is the role of operations research analysts?
Operations research analysts are professionals who apply advanced analytical methods to help organizations investigate complex issues, identify and solve problems, and make better decisions. Some key responsibilities include:
- Building mathematical models to represent business operations or processes
- Gathering input data for analysis and identifying relevant variables
- Applying algorithms and optimization techniques to identify optimal solutions
- Analyzing tradeoffs of alternative solutions
- Performing sensitivity analysis to evaluate the effects of different parameters
- Using simulation and risk analysis to quantify uncertainty
- Developing optimization software and decision-support systems
- Communicating analytical insights to stakeholders using data visualization
- Continuously monitoring solutions and models to adapt as conditions change
Operations research analysts are highly skilled professionals. They require training in operations research, applied mathematics, statistics, analytics, modeling, and computer science. Analysts need strong critical thinking skills to translate business issues into quantitative models. Communication skills are necessary to relay findings to decision makers. Operations research roles are found at companies in many industries, management consulting firms, and government agencies.
How does operations research relate to data science and machine learning?
There is significant overlap between operations research and the newer disciplines of data science and machine learning. All three fields leverage statistical modeling, algorithms, and computational analysis to find patterns and insights in data. However, some key differences include:
| Operations Research | Data Science | Machine Learning |
|---|---|---|
| Focuses on mathematical optimization of operational processes | Extracts insights from large, unstructured datasets | Develops algorithms that can learn and improve autonomously |
| Prescriptive analytics | Descriptive and predictive analytics | Predictive analytics and pattern recognition |
| Knowledge of operations is key | Knowledge of statistics and modeling is key | Knowledge of computer science is key |
There is significant overlap between the fields, and analysts may leverage techniques from all three. But operations research maintains a distinct focus on mathematical optimization and decision science for operational systems.
What are common criticisms of operations research?
Some common criticisms leveled at the field of operations research include:
- Over-reliance on mathematical models that may not capture real-world complexity
- Questionable validity of model assumptions
- Difficulty obtaining accurate and complete input data
- Solutions may be theoretical rather than practical
- Lagging behind latest technological advances in analytics and computing
- Fails to consider “soft” variables like human behavior and politics
- Too academic and isolated from actual operations
While these criticisms may have some validity, they can often be addressed by skilled analysts. The best operations research combines mathematical rigor with practical judgment and experience. Analysts must make reasonable simplifying assumptions and approximations. Solutions must be stress-tested for robustness against incomplete data and model uncertainty. Organizations get the most value when operations research teams are fully integrated with operations staff and decision makers.
Has operations research declined in usage and importance?
While early operations research mainly focused on mathematical optimization for military and industrial operations, the field has expanded significantly over the past few decades. Operations research is now applied across sectors like transportation, energy, healthcare, retail, and finance. New opportunities have opened up with the availability of large datasets and cheap computing power.
That said, operations research has faced competition from upstart fields like machine learning that have captured more attention and talent. Enrollments in operations research academic programs have declined since the 1980s. But this may be due less to operations research becoming obsolete, and more to students being drawn into related disciplines under the banner of analytics, data science, and AI.
Powerful new optimization solvers like Gurobi and CPLEX allow practitioners to solve ever larger problems. Operations research enables organizations to unearth insights that other methodologies may miss. The field continues to develop innovative new methodologies like agent-based modeling. Claims of the demise of operations research seem unfounded.
How has operations research evolved and innovated?
Operations research has evolved significantly since World War II, expanding into many application areas. Some key innovations include:
- Faster computers – Allowing solution of larger, complex models with many variables and constraints.
- New algorithms – Novel optimization techniques like genetic algorithms and ant colony optimization inspired by natural processes.
- Simulation – Computer-based simulation provides a versatile tool for analyzing stochastic, dynamic systems.
- Big data analytics – OR methods being adapted for complex analytics on high-volume, unstructured data.
- Machine learning integration – OR and ML complement each other’s strengths.
- Decision science – Understanding how psychology and organizational behavior impacts decision making.
- Behavioral OR – Accounting for human cognitive biases and risk preferences.
Rather than becoming outdated, operations research continues to adopt the latest analytical innovations, helping decision makers solve increasingly complex problems. The field retains its essential character while steadily evolving.
What new applications are emerging for operations research?
Exciting new applications for operations research include:
- Healthcare – Optimizing patient flow, surgery scheduling, hospital layouts, ambulance dispatching.
- Ride sharing – Vehicle assignment optimization for Uber, Lyft, etc.
- Robotic warehouses – Coordinating autonomous warehouse robots and maximizing efficiency.
- Drone delivery – Route optimization and coordination for emerging drone delivery services.
- Renewable energy – Designing and integrating solar, wind and other renewable generation.
- Smart cities – Optimizing traffic flows, transit systems, infrastructure management.
- Defense – Battle planning, logistics, cybersecurity.
As technology progresses and systems become more complex, operations research will continue to find new vital applications across many industries.
How is operations research used in business today?
Modern businesses use operations research in many ways, including:
- Supply chain optimization – Strategic distribution network design, transportation mode selection, production planning.
- Scheduling – Production schedules, airline and train crew schedules, call center staffing.
- Route planning – Delivery truck route optimization, coordinating installation/repair crews.
- Pricing optimization – Leveraging operations data to optimize pricing.
- Quality management – Statistical quality control techniques.
- Layout optimization – Optimizing factory or warehouse layouts.
- Inventory management – Optimizing inventory levels across supply chains.
- Maintenance optimization – Scheduling maintenance to minimize disruptions.
The analytical tools of operations research allow businesses to maximize productivity and efficiency in complex systems with many variables. OR is extensively used in large corporations across all industries.
What skills are required to be a modern operations research practitioner?
To be an effective operations research analyst today requires a cross-disciplinary skillset, including:
- Mathematics – Linear algebra, calculus, statistics, algorithms.
- Computer Science – Programming, modeling, database skills.
- Analytics – Descriptive, predictive and prescriptive analytics methodologies.
- Data Visualization – Communicating data insights effectively.
- Scientific Method – Hypothesis testing, experimental design.
- Critical Thinking – Problem structuring, analytical reasoning.
- Communication – Translating analytical findings into actionable business insights.
Operations research analysts must balance technical analytical skills with business acumen and communication ability. Ongoing education is needed as the field rapidly evolves.
What is the future outlook for operations research?
Operations research will continue to play a vital role in the increasingly complex systems and data-rich environments of the future. Like other analytical disciplines, operations research will need to keep adopting the latest advances in data science, statistics, and machine learning. To remain relevant, the field needs to attract top talent and focus on high-value applications.
With omnipresent data and cheap computing power, organizations require optimization capabilities more than ever. While some analytical tasks are being automated, human operations research practitioners are still needed to frame problems, interpret results, and make dynamic decisions. The abundance of data also allows more sophisticated models that better reflect real-world complexities.
Operations research will be vital for innovations like smart cities, autonomous vehicles, renewable energy, and next-gen healthcare. The field must continue evolving, while retaining its identity and expanding awareness of its capabilities.
Conclusion
While some have predicted the demise of operations research in the face of new analytical approaches, the field remains highly relevant. OR provides unique prescriptive capabilities to model complex systems and derive optimal solutions. It continues to find new applications in emergent industries.
Like any longstanding discipline, operations research needs to adapt to remain vibrant. New methodologies and computing capabilities must be embraced. A focus on high-impact applications and communication of value to business stakeholders is critical. With these adjustments, operations research is poised to keep benefiting organizations and society for decades to come.
