Tag: Research Tools

The Role of Natural Language Processing in Assessing Demand for New Infrastructure

 In the dynamic landscape of urban development and planning, the assessment of demand for new infrastructure is pivotal. As cities grow and evolve, the efficient planning of infrastructure—from roads and bridges to public transport and utilities—is crucial for sustainability and quality of life. Enter Natural Language Processing (NLP), a branch of artificial intelligence that has the potential to significantly enhance the methodologies used in infrastructure demand assessment.

Understanding NLP and Its Capabilities

Natural Language Processing involves the interaction between computers and humans through natural language. The goal of NLP is to read, decipher, understand, and make sense of human languages in a manner that is valuable. This technology processes large amounts of natural language data to extract insights and patterns that are not readily apparent to humans.

NLP in Infrastructure Demand Assessment

1. Data Collection and Analysis

  • Social Media and Online Forums: NLP can analyze discussions and sentiments expressed on social media platforms and online forums regarding infrastructure needs. By examining tweets, posts, and comments, NLP tools can gauge public opinion on existing infrastructure and potential demand for new projects.
  • Survey Data: Traditional surveys generate vast amounts of textual data, often in the form of open-ended responses. NLP can automate the analysis of these responses, providing quick and detailed insights into public sentiment and demand.

2. Predictive Analytics

  • Trend Analysis: NLP can identify trends in public opinion and emerging needs by analyzing changes in language and topics over time. This helps in predicting future demands and potential infrastructure challenges.
  • Sentiment Analysis: By assessing the sentiment behind the textual data gathered from various sources, NLP helps in understanding the public’s feelings towards proposed or existing infrastructure projects.

3. Enhancing Public Engagement

  • Feedback Mechanisms: NLP can be used to develop smarter feedback mechanisms where public input on infrastructure projects is gathered and analyzed in real-time. This can significantly improve the responsiveness of planning authorities to public needs.
  • Chatbots and Virtual Assistants: These tools can be deployed to interact with the public, gather data, and provide information on infrastructure projects, enhancing engagement and participation.

Case Studies and Applications

  • Singapore’s Smart Nation Initiative: Utilizing NLP to analyze communications and feedback on urban planning efforts, helping to guide decisions on where to focus infrastructural developments.
  • Transport for London (TfL): Using sentiment analysis to assess public reactions to changes in transport services and infrastructure, aiding in better decision-making and service adjustments.

Challenges and Considerations

While NLP offers substantial benefits, there are challenges in its implementation:

  • Data Privacy and Security: Handling large volumes of personal data requires robust security measures and adherence to privacy laws.
  • Accuracy and Context: NLP algorithms must be finely tuned to accurately interpret the nuances and context of language, which varies widely across different cultures and communities.

The Future of NLP in Urban Planning

As NLP technology advances, its integration into urban planning and infrastructure development is expected to deepen. Future applications could involve more advanced predictive models and real-time public sentiment analysis, leading to more responsive and effective urban infrastructure planning.

NLP presents a transformative potential for urban planning, offering a more nuanced understanding of public needs and expectations. By harnessing this technology, planners and policymakers can improve the efficiency and efficacy of infrastructure development, ultimately leading to smarter, more sustainable cities.

Using Modeling and Simulation to Test Designs and Requirements

 Modeling is an efficient and cost-effective way to represent a real-world system. A model can represent key aspects of the system, including the underlying requirements, the components, and how those components communicate with one another. The model can be simulated, enabling designers to test designs before hardware is available, or to test conditions that are either difficult or expensive to replicate in the real world. Iterating between modeling and simulation can improve the quality of the system design early, reducing the number of errors found later in the design process.

Despite these advantages, designers who heavily rely on hand coding do not always take full advantage of modeling and simulation. Setting up tests can be difficult and time-consuming, and when separate tools are used for each domain, it can be challenging to obtain a system-level view of the design. As a result, defects that could have been found in the modeling and simulation phase are often found during the implementation phase, when defects are more expensive to fix.

These issues are addressed in Simulink®, a platform for modeling and simulation. Simulink supports not only multidomain modeling but also simulation, with its own set of ordinary differential equation (ODE) solvers. A fundamental advantage of using Simulink is that you can represent different domains, including control systems, state machines, and environmental models, in one model, and then run simulations within Simulink to verify that the model is built correctly. As the simulation runs, you have access to simulation analysis capabilities, such as data displays, state animation, and conditional breakpoints. After the simulation is completed, you can analyze any logged data with MATLAB® scripts and visualization tools.

In this article, we describe a workflow for building a component model from requirements, simulating and testing that component model, and then connecting it to a system-level model for further simulation and testing. To illustrate this workflow we will build and test the fault detection, isolation, and recovery (FDIR) component of the HL-20, a re-entry vehicle designed by NASA to complement the Space Shuttle orbiter. We will connect our component to a system-level model that includes environmental models, flight controls, and guidance, navigation, and controls (GN&C) systems, and then simulate the system-level model to validate its behavior.

What is Simulation Modelling

SN Sharma (2022) has rightly stated that “Simulation is a well designed model which should be representative of real world like situation, it can be done after taking into account all components with well-defined (cohesive) functions, and minimised dependencies on other component which can be done after getting thorough knowledge of the functioning conditions of the model.”

If we restrict ourselves to simulation software, and to simulation design (ignoring aspects of development process and code access), there are three main strongly-related areas of best-practice which I would regard as universal to all simulation (precisely because they are universal to all software), and which are echoed in computational science best-practice papers (Sandve et al. 2013; Wilson et al. 2014), software engineering textbooks (Sommerville 2011), and practitioner best-practice handbooks (McConnell 2004); all backed by empirical research (Oram & Wilson 2010):

Automated Reproducibility
Being able to recreate any run of the software—for testing purposes and to check claims about its outputs—in an automated way (not just via manual recreation from documentation). This includes provenance (and perhaps automated recreation) of the entire computational environment (since results can vary based on things like the versions of external libraries used).

Cohesive, Loosely-Coupled Design
A design separated into components with well-defined (cohesive) functions, and minimised dependencies on other components (loose-coupling). This massively aids the debugging, maintenance and reusability of the code. This often involves reusing recurring structural and behavioural forms that have been shown to help solve common design issues: SE calls such forms design patterns (Gamma et al. 1995; Buschmann 1996). Such forms help establish a shared software design vocabulary at a higher level of abstraction.

Testability
Being designed in a way that facilitates testing at different levels (e.g., single class, component or whole system) and, where possible, includes automated tests as part of the software deliverable. In particular, automated tests provide a bank of regression tests which can be continually re-run to check that changes have not caused bugs elsewhere (i.e., caused previously successful tests to fail). Such tests become the central driver of the development process in the increasingly-used Test-Driven Development (TDD) approach (Jeffries & Melnik 2007).

1.3
Software exhibiting these properties is highly reusable (given access to it) and its implementation will typically involve reuse of existing software where it exists. In the simulation domain, there are many toolkits which provide reusable, well-tested software for simulation development which include (a) templates for model elements relating to one or more modelling paradigms—such as agent-based modelling (ABM), discrete-event simulation (DES) or system dynamics (SD); and (b) supporting infrastructure code to create and run models, such as for visualisations, simulation control interfaces, and input/output handling.

1.
There has been some emerging work which tries to define new simulation frameworks and abstractions which better embody some of these principles; e.g., the modular architecture and best-practice of JAMES II (Uhrmacher 2012; Himmelspach & Uhrmacher 2007), or test and experiment specifications which are model-based (Djanatliev et al. 2011) or domain-language-based (Ewald & Uhrmacher 2014).

However, there appears to be virtually no discussion of these issues more generally for ‘mainstream’ simulation using widely-used toolkits such as, in the ABM case, NetLogo (Tisue & Wilensky 2004), Repast Simphony (North et al. 2013), MASON (Luke et al. 2005), or AnyLogic (Borshchev & Filippov 2004). In particular, there is nothing which allows simulation practitioners to understand how these ideas might be embodied in some best-practice simulation design, and to therefore have some frame to assess existing toolkits and make more informed decisions on their choice of simulation platform (and thus understand the strengths and weaknesses of their simulation software design with respect to this best-practice).

Utilities.
General utilities (not specific to the domain model or upper layers) for (a) data types (e.g., linked lists); (b) input/output capabilities, such as to/from different file formats; and (c) general algorithmic facilities such as random number generators, probability distributions or differential equation numerical solvers. These can interact; e.g., probability distributions could be initialised from external files.

Domain Model.
The code representing the abstraction of the real-world system, including the representation of space and time.

Execution Control.
How the domain model is actually executed, which typically amounts to instantiating a ‘root’ object and stepping through a schedule of actions (provided by a domain model component) to ‘unfold’ time dynamically. Because non-domain-model objects also need to interleave their actions in simulated time, this layer includes that capability. This is a ‘thin’ layer, but nevertheless a well-defined one.

Meta-Data Capture.
Code (scheduled in simulation time) to capture and calculate meta-data; i.e., derived model state (possibly held as a time series to capture changes over time) or atomic model state captured over time.

This layer includes any writing of outputs to file (or database) because this can be tightly coupled with meta-data capture; in larger-scale simulations, time series data may be captured in a rolling window for storage reasons (perhaps with this window used for visualisation), with outputs written to file as they ‘drop out of’ the window (or via some other buffering strategy).

State & Control Presentation.
The parts of the user interface which present model state and controls as part of a user interface. The presentation may be visual or textual. Where current model state is being presented, this directly uses the Domain Model layer. If meta-data is being presented, this uses the Meta-Data Capture layer.

In particular, note that a given domain model might have multiple presentations, with multiple alternative visualisations per component; such solutions require a layered domain model separation.

Experiments Definition.
The parts of the user interface which support the definition of simulation runs (experiments), possibly including multi-run experiments. This mainly consists of how model inputs are defined and passed on to the model, and any automated manipulation of them across multiple runs for things like sensitivity analysis. Because this tends to be particularly generic to any simulation (and modellers using multiple toolkits may want a vendor-neutral solution), separate experimental platforms exist (Gulyás et al. 2011), and I am aware of simulation consultancies who develop their own in-house.

References

 

ALLAN, R. J. (2010). Survey of agent-based modelling and simulation tools. Tech. Rep. DL-TR-2010-007, Science & Technologies Facilities Council (STFC), UK.

BORSHCHEV, A. & Filippov, A. (2004). From system dynamics and discrete event to practical agent based modeling: Reasons, techniques, tools. In: Proceedings of the 22nd International Conference of the System Dynamics Society.

BRAILSFORD, S. C. (2014). Discrete-event simulation is alive and kicking! Journal of Simulation 8(1), 1–8.

BUSCHMANN, F. (1996). Pattern-Oriented Software Architecture: a system of patterns. Wiley, volume 1 ed. http://www.worldcat.org/isbn/0471958697.

COLLIER, N. & Ozik, J. (2013). Test-driven agent-based simulation development. In: Proceedings of the 2013 Winter Simulation Conference (Pasupathy, R., Kim, S. H., Tolk, A., Hill, R. & Kuhl, M. E., eds.). IEEE.

DAVISON, A. P., Mattioni, M., Samarkanov, D. & Teleńczuk, B. (2014). Sumatra: A toolkit for reproducible research. In: Implementing Reproducible Research (Stodden, V., Leisch, F. & Peng, R. D., eds.), chap. 3. Chapman & Hall, pp. 57–78.

DJANATLIEV, A., Dulz, W., German, R. & Schneider, V. (2011). VERITAS—a versatile modeling environment for test-driven agile simulation. In: Proceedings of Wintersim 2011 (Jain, S., Creasey, R. R., Himmelspach, J., White, K. P. & Fu, M., eds.). http://www.informs-sim.org/wsc11papers/325.pdf.

EVANS, E. (2004). Domain-Driven Design: tackling complexity in the heart of software. Addison-Wesley.

EWALD, R. & Uhrmacher, A. M. (2014). SESSL: A domain-specific language for simulation experiments. ACM Trans. Model. Comput. Simul. 24(2). . [doi:10.1145/2567895]

FOWLER, M. (2004). UML Distilled: A Brief Guide to the Standard Object Modeling Language. Addison-Wesley, third ed. http://www.worldcat.org/isbn/0321193687.

FOWLER, M., Rice, D., Foemmel, M., Hieatt, E., Mee, R. & Stafford, R. (2003). Patterns of Enterprise Application Architecture. Addison-Wesley.

GAMMA, E., Helm, R., Johnson, R. & Vlissides, J. (1995). Design Patterns: Elements of Reusable Object-Oriented Software. Addison-Wesley.

GILBERT, N. & Troitzsch, K. G. (2005). Simulation for the Social Scientist. Open University Press, 2nd ed.

GREEN, T. R. G. & Petre, M. (1996). Usability analysis of visual programming environments: A ‘cognitive dimensions’ framework. Journal of Visual Languages & Computing 7(2), 131–174. . [doi:10.1006/jvlc.1996.0009]

GRIMM, V. & Railsback, S. F. (2005). Individual-based modeling and ecology. Princeton Series in Theoretical and Computational Biology. Princeton University Press. http://www.worldcat.org/isbn/0691096651.

GULYÁS, L., Szabó, A., Legéndi, R., Máhr, T., Bocsi, R. & Kampis, G. (2011). Tools for large scale (distributed) agent-based computational experiments. In: Proceedings of CSSSA 2011. Computational Social Science Society of the Americas (CSSSA).

GÜRCAN, O., Dikenelli, O. & Bernon, C. (2013). A generic testing framework for agent-based simulation models. Journal of Simulation 7(3), 183–201. . [doi:10.1057/jos.2012.26]

HIMMELSPACH, J. & Uhrmacher, A. M. (2007). Plug’n simulate. In: 40th Annual Simulation Symposium (ANSS’07). Washington, DC, USA: IEEE. .

JEFFRIES, R. & Melnik, G. (2007). TDD: The art of fearless programming. IEEE Software 24(3), 24–30. . [doi:10.1109/MS.2007.75]

JOINES, J. A. & Roberts, S. D. (1999). Simulation in an object-oriented world. In: Proceedings of the 1999 Winter Simulation Conference (Farrington, P. A., Nembhard, H. B., Sturrock, D. T. & Evans, G. W., eds.). .

KERNIGHAN, B. (1979). UNIX for Beginners. Bell Telephone Labs, 2nd ed.

LUKE, S., Cioffi-Revilla, C., Panait, L., Sullivan, K. & Balan, G. (2005). MASON: A multiagent simulation environment. Simulation 81(7), 517-527. [doi:10.1177/0037549705058073]

MCCONNELL, S. (2004). Code Complete: A practical handbook of software construction. Microsoft Press, second ed. http://www.worldcat.org/isbn/9780735619678.

MILLER, J. H. & Page, S. E. (2007). Complex Adaptive Systems: An Introduction to Computational Models of Social Life. Princeton Studies in Complexity. Princeton Press.

MILLINGTON, J. D. A., O’Sullivan, D. & Perry, G. L. W. (2012). Model histories: Narrative explanation in generative simulation modelling. Geoforum 43(6), 1025–1034. . [doi:10.1016/j.geoforum.2012.06.017]


MULDER, J. D., van Wijk, J. J. & van Liere, R. (1999). A survey of computational steering environments. Future Generation Computer Systems 15(1), 119–129. . [doi:10.1016/S0167-739X(98)00047-8]

MÜLLER, J. P. (2009). Towards a formal semantics of event-based multi-agent simulations. In: Multi-agent Based Simulation IX, no. 5269 in LNCS. Springer.

NIKOLAI, C. & Madey, G. (2009). Tools of the trade: A survey of various agent based modeling platforms. Journal of Artificial Societies and Social Simulation, 12(2), 2. https://www.jasss.org/12/2/2.html.

NORTH, M. J., Collier, N. T., Ozik, J., Tatara, E. R., Macal, C. M., Bragen, M. & Sydelko, P. (2013). Complex adaptive systems modeling with Repast Simphony. Complex Adaptive Systems Modeling 1(1), 3+. . [doi:10.1186/2194-3206-1-3]

NORTH, M. J. & Macal, C. M. (2014). Product and process patterns for agent-based modelling and simulation. Journal of Simulation, 8, 25-36. . [doi:10.1057/jos.2013.4]

ORAM, A. & Wilson, G. (eds.) (2010). Making software : what really works, and why we believe it. O’Reilly. http://www.worldcat.org/isbn/9780596808327.

RAILSBACK, S. F. & Grimm, V. (2012). Agent-based and individual-based modeling : a practical introduction. Princeton University Press. http://www.worldcat.org/isbn/9780691136745.

RAILSBACK, S. F., Lytinen, S. L. & Jackson, S. K. (2006). Agent-based simulation platforms: review and development recommendations. Simulation 82, 609–623. http://www.humboldt.edu/ecomodel/documents/ABMPlatformReview.pdf. [doi:10.1177/0037549706073695]

ROPELLA, G. E., Railsback, S. F. & Jackson, S. K. (2002). Software engineering considerations for individual-based models. Natural Resource Modeling 15(1), 5–22. [doi:10.1111/j.1939-7445.2002.tb00077.x]

ROUCHIER, J., Cioffi-Revilla, C., Polhill, J. G. & Takadama, K. (2008). Progress in model-to-model analysis. Journal of Artificial Societies and Social Simulation 11(2), 8. https://www.jasss.org/11/2/8.html.

SANDVE, G. K., Nekrutenko, A., Taylor, J. & Hovig, E. (2013). Ten simple rules for reproducible computational research. PLoS Comput Biol 9(10), e1003285+. . [doi:10.1371/journal.pcbi.1003285]

SEGAL, J. (2008). Scientists and software engineers: A tale of two cultures. In: Proceedings of the Psychology of Programming Interest Group PPIG 08.

SOMMERVILLE, I. (2011). Software engineering. Pearson, 9th ed. http://www.worldcat.org/isbn/9780137053469.

STODDEN, V., Donoho, D., Fomel, S., Freidlander, M. P., Gerstein, M., Leveque, R., Mitchell, I., Larrimore Ouellette, L. & Wiggins, C. (2010). Reproducible research: Addressing the need for data and code sharing in computational science. Computing in Science & Engineering 12(5), 8–13. .

STODDEN, V., Guo, P. & Ma, Z. (2013). Toward reproducible computational research: An empirical analysis of data and code policy adoption by journals. PLoS ONE 8(6), e67111+. . [doi:10.1371/journal.pone.0067111]

TISUE, S. & Wilensky, U. (2004). NetLogo: Design and implementation of a multi-agent modeling environment. In: Proceedings of Agent 2004.

UHRMACHER, A. M. (2012). Seven pitfalls in modeling and simulation research. In: Proceedings of the 2012 Winter Simulation Conference (Laroque, C., Himmelspach, J., Pasupathy, R., Rose, O. & Uhrmacher, A. M., eds.).

VIANA, J., Rossiter, S., Channon, A. A., Brailsford, S. C. & Lotery, A. (2012). A multi-paradigm, whole system view of health and social care for age-related macular degeneration. In: Proceedings of the Winter Simulation Conference, WSC ’12. Winter Simulation Conference. http://dl.acm.org/citation.cfm?id=2429759.2429884.

WHITLEY & Blackwell, A. F. (2001). Visual programming in the wild: A survey of LabVIEW programmers. Journal of Visual Languages & Computing 12(4), 435–472. . [doi:10.1006/jvlc.2000.0198]

WILSON, G. (2014). Software carpentry: lessons learned. F1000Research . [doi:10.12688/f1000research.3-62.v1]

WILSON, G., Aruliah, D. A., Brown, C. T., Chue Hong, N. P., Davis, M., Guy, R. T., Haddock, S. H. D., Huff, K. D., Mitchell, I. M., Plumbley, M. D., Waugh, B., White, E. P. & Wilson, P. (2014). Best practices for scientific computing. PLoS Biol 12(1), e1001745+. . [doi:10.1371/journal.pbio.1001745]

ZEIGLER, B. P., Gon Kim, T. & Praehofer, H. (2000). Theory of modeling and simulation : integrating discrete event and continuous complex dynamic systems. Academic Press, 2nd ed.

ZINN, S., Himmelspach, J., Uhrmacher, A. M. & Gampe, J. (2013). Building Mic-Core, a specialized M&S software to simulate multi-state demographic micro models, based on JAMES II, a general M&S framework. Journal of Artificial Societies and Social Simulation 16(3), 5. https://www.jasss.org/16/3/5.html.

https://www.mathworks.com/company/newsletters/articles/using-modeling-and-simulation-to-test-designs-and-requirements.html