Skip to content
TheChatArchiveThe Chat Archive
gpt-4-0314· OpenAIAi2 WildChat-4.8M (AI2)

As a Masters degree student, YOu Have a project and the brief goes thus: Component A is…

  1. User

    As a Masters degree student, YOu Have a project and the brief goes thus:

    Component A is a team presentation of 25 minutes (with 5 minutes of questions).

    As a group you will be required to create a PowerPoint presentation which will detail the planning and management issues of a global project.

    1.Choose a project you know off or you are interested in provided it is global, whether completed or in progress, and make sure that you are able to gather the information that is necessary for your analysis;

    2.There are two sets of slides in Blackboard > Learning Material > week 1. These slides give you directions how to use UWE resources to find a case project.

    3.Use the tools as described below in section 3. Content below and participate in the workshops to practice each tool with your group.

    The presentation will be live, and all group members will present ONE part of the Assignment for which they will be individually marked. Not all group members will get the same mark necessarily, it depends on the quality of their work and presentation in their own part. Group members who do not participate and contribute to the Assignment will have to re-sit as they will get 0. .

    This assessment strategy is employed to produce the type of output required by professional project management practice that is developed by a typical project team; and to maximize learning opportunities with added values for students’ development and character building.

    This assignment requires you to develop research skills and find information that is useful for the purpose of detecting organizational issues related to the global environment and construct a plan. You will have to work with University databases and search through a large volume of information from Professional associations.

    If you have difficulty to choose a global project here are some choices:

    SMART CITIES EXAMPLES https://www.apm.org.uk/blog/five-of-the-uk-s-most-exciting-smart-city-projects/ https://publications.iadb.org/en/international-case-studies-smart-cities-santander-spain

    GLOBAL HEALTHCARE PROJECTS EXAMPLES https://www.bond.org.uk/news/2017/04/7-projects-improving-healthcare-around-the-world

    GLOBAL MEDICAL AND BIOTECH RESEARCH PROJECTS EXAMPLES https://globalprojects.ucsf.edu/ https://www.nihr.ac.uk/explore-nihr/funding-programmes/global-health.htm https://www.tendersinfo.com/healthcare-and-medical-pharma-projects.php https://www.pharmaceuticalonline.com/doc/top-10-international-pharmaceutical-biotech-c-0001

    GLOBAL HUMANITARIAN PROJECTS EXAMPLES https://borgenproject.org/5-top-humanitarian-aid-organizations/ https://www.humanitarianprojectsinc.com/

    These are some examples, you can find global projects in most industries.

    Please do not just use Apple, Google or Toyota or any of these famous large organizations because you cannot have access to information about their projects. Choose projects about which you can discover information.

    I strongly suggest that you attend the workshops and be proactive in forming groups in week 1 and use the Direction files in the Learning Materials folders showing you case study books to select project case studies.

    2.Presentation Content & Structure The live team presentations should not exceed 30 minutes including Q&A and feedback from the tutor. This means that the presentation should be approximately 25 minutes maximum. In a team of 5 people this means that each person will have to present their contribution for 5 minutes.

    It is the responsibility of the team to allocate the parts of the work amongst them.

    Content

    Your task is to present your analysis & plan of a global project. You will choose the project as described in the section above.

    STRUCTURE OF THE PRESENTATION - PARTS:

    INTRODUCTION: introduce your project and your team and each team member’s contribution you are going to have a slide where you will attach next to each part the name of the team member who worked at it and who will present it.

    ANALYSIS PART 1. Evaluate the challenges regarding such a project using the dimensions model from Binder
    Evaluate the impact of environmental factors on this project. oCollect information about the project environment through industry & professional reports and company literature or case studies. You do not need to do primary research and collect information from the organization people.

    PLAN PART 2: Recommend the most appropriate method for managing your project using the Global Project Management Framework from Binder. This part will focus on: 1.a risk assessment based on the Global Project Management Framework 2.explain how you will deal with the issues of: international and intercultural communications, global team structures, and collaborative relationships with other organizations in the particular project 3.The team and project manager roles & skillset [competency maps/leadership framework/Belbin roles]

    PLAN PART 3: Construct a Project Plan with objectives, a phase plan and collaboration tools and collaboration techniques as advised by Binder *Note: be careful that the type of plan you present is linked to the methodology you chose. For example, if you select a specific agile methodology your plan should be based on the type of iterations of this method.

    References: reports from industry or from organization/firm, company material and information from reputable sources (case studies, professional magazines articles etc.) and the sources from the models you use to conduct analysis

    3.Preparation for Assignment A (mandatory)

    Students are advised to attend the workshops because the Assignment consists of the use of a series of tools, each tool belonging to a part of the Assignment. Please look at the table below to see which workshops correspond to which tools and which parts of the Assignment.

    Workshops Section in Assignment A Tool to use Workshop Session Find a project Introduction Use the slides Session 1 Detect the challenges Part 1 The Radar Chart Session 2 Detect success factors Part 1 The PIP Session 3 Describe the team competencies needed Part 2 Competency map Session 4 Design the project manager role & leadership Part 2 leadership framework Session 5 Describe the teams’ behavioural roles Part 2 Belbin team roles Session 6 Choose the way your teams will be creative Part 2 Brainstorming Techniques Session 7 Describe the cultural differences and challenges in your project Part 2 Hofstede cultural scores Session 8 Explain how you qill uaw technologies to deal with challenges Part 3 Collaboration platforms Session 9 Design project plan for your presentation Part 3 Already known tools Session 10

    In preparing for this assessment, teams will be able to undertake a series of formative exercises and gain feedback in response. Expectations concerning the presentation will be discussed in workshops, and teams will also be able to present ONE draft presentation plan for formative review on the final workshop on week 10.

    As detailed as possible, explain step by step how to get this done

  2. Assistant

    Step 1: Form a group and select the project

    • Gather a group of classmates (ideally 4-5 people).
    • Select a global project you all are interested in and make sure there is enough information available for your analysis.
    • Consider the project options provided in the brief or find your own global project.

    Step 2: Assign roles

    • Allocate different parts of the presentation to each team member.
    • Make sure each team member has roughly equal contributions.

    Step 3: Attend workshops

    • Attend mandatory workshops listed in the timetable provided in the brief.
    • Use the tools learned in the workshops for different sections of the presentation.

    Step 4: Research and collect data

    • Search through university databases and professional associations for relevant information.
    • Look for organizational and environmental issues related to your chosen global project.
    • Keep track of your sources for citation purposes.

    Step 5: Create the presentation

    • Work together to create a PowerPoint presentation that includes the following sections:

      A. Introduction: Introduce the project, the team, and individual contributions.

      B. Analysis Part 1: Evaluate challenges related to the project using the dimensions model from Binder. Assess the impact of environmental factors on the project using data collected from industry reports, company literature, or case studies.

      C. Plan Part 2: Recommend a project management method using the Global Project Management Framework from Binder. Focus on risk assessment, international and intercultural communication, global team structures, collaborative relationships, and the roles and skillsets of the project manager and team.

      D. Plan Part 3: Construct a project plan with objectives, a phase plan, and collaboration tools and techniques based on Binder's suggestions. Ensure consistency with the methodology chosen in Plan Part 2.

    Step 6: Practice the presentation

    • Ensure the presentation does not exceed 25 minutes.
    • Practice individual contributions and transitions between sections.
    • Allocate equal times (approximately 5 minutes) for each team member's presentation.

    Step 7: Utilize feedback from formative exercises

    • Engage in formative exercises and seek feedback on your presentation.
    • Make necessary adjustments based on feedback from the instructor or peers.

    Step 8: Attend and present at the final workshop

    • Present your analysis and plan for the chosen global project on the final presentation day.
    • Ensure the live team presentation does not exceed 30 minutes, including Q&A and feedback from the tutor.
    • Each group member should present their contribution individually.

    Step 9: Address questions and receive feedback

    • During the Q&A session following the presentation, be prepared to address any questions or concerns raised by the tutor or classmates.
    • Take note of any feedback provided and consider incorporating it into future projects.
  3. User

    Your part of the project is to do the sessions 7 & 8; that is, Choose the way your teams will be creative (Brainstorming Techniques) and Describe the cultural differences and challenges in your project (Hofstede cultural scores) How would you get this done?

  4. Assistant

    To complete sessions 7 and 8, follow the steps outlined below:

    Step 1: Understand the tools

    • Familiarize yourself with Creative Brainstorming Techniques and Hofstede's Cultural Dimensions.
    • Research and learn how these tools work and their applications in a global project context.

    Step 2: Attend the workshops

    • Attend the required workshops for sessions 7 and 8.
    • Engage with the workshop activities and discussions to gain a deeper understanding of the tools.

    Step 3: Apply Brainstorming Techniques to your project

    • Work with your team members to select appropriate brainstorming techniques for enhancing creativity and problem-solving in your project.
    • Some techniques to consider include mind-mapping, round-robin brainstorming, the six thinking hats, brainwriting, and the reverse brainstorming method.
    • Discuss with your team how you would apply these techniques in different stages of the project.

    Step 4: Analyze cultural differences using Hofstede's Cultural Dimensions

    • Research the countries involved in the project and gather data on their cultural dimensions based on Hofstede's model.

    • Hofstede's six cultural dimensions include:

      1. Power Distance Index (PDI)
      2. Individualism vs. Collectivism (IDV)
      3. Masculinity vs. Femininity (MAS)
      4. Uncertainty Avoidance Index (UAI)
      5. Long-term Orientation vs. Short-term Orientation (LTO)
      6. Indulgence vs. Restraint (IVR)
    • Compare and contrast the cultural dimensions of the countries involved in your project.

    Step 5: Identify cultural challenges and opportunities

    • Analyze the differences and similarities among the countries in the project based on Hofstede's cultural dimensions.
    • Identify potential cultural challenges, conflicts, or misunderstandings that may arise due to these differences.
    • Consider opportunities for cultural synergy, learning, and collaboration among team members from different cultural backgrounds.

    Step 6: Integrate your findings with the team presentation

    • Prepare your section of the PowerPoint presentation based on your analysis and findings.
    • Make sure your presentation aligns with the contributions of your team members from other workshop sessions.
    • Ensure your part of the presentation covers both brainstorming techniques and the cultural differences and challenges analysis.

    Step 7: Coordinate with your group

    • Share your findings with your team members and solicit feedback from them.
    • Discuss how your findings might impact other parts of the project, such as communication strategies, team structure, and project management approaches.
    • Make necessary adjustments based on group discussions and feedback.

    Step 8: Practice your part of the presentation

    • Rehearse your section of the presentation with the rest of your team.
    • Allocate your speaking time to approximately 5 minutes, focusing on the essential aspects of your analysis.
    • Ensure your presentation is clear, concise, and engages the audience in understanding the creative brainstorming techniques and cultural aspects of the project.
  5. User

    In the best possible way, give me a structure for my ppt presentation. Break it down into slides

  6. Assistant

    Here is a suggested structure for your PowerPoint presentation with a breakdown of slides:

    Slide 1: Title Slide

    • Project title
    • Team member names
    • Date

    Slide 2: Introduction

    • Brief project overview
    • Objective of the presentation
    • Outline of presentation sections and team member responsibilities

    Slide 3: Project Background

    • Brief description of the chosen global project
    • Problem statement / goal of the project
    • Countries or locations involved

    Slide 4: Brainstorming Techniques - Introduction

    • Importance of creative brainstorming in global projects
    • Brief overview of selected brainstorming techniques

    Slides 5-9: Brainstorming Techniques – Details

    • Explanation of each selected brainstorming technique (one technique per slide)
    • Include examples, benefits, and a brief explanation of how the technique will be applied in the project

    Slide 10: Cultural Differences - Introduction

    • Importance of understanding cultural differences in global projects
    • Brief introduction to Hofstede's Cultural Dimensions

    Slides 11-16: Hofstede's Cultural Dimensions - Country Comparisons

    • A slide for each cultural dimension, comparing the countries involved in the project
    • Include graphs or charts to illustrate differences and similarities

    Slide 17: Cultural Challenges and Opportunities

    • Summarize potential cultural challenges based on the cultural dimensions analysis
    • Suggest strategies to address these challenges
    • Highlight opportunities for cultural synergy and learning

    Slide 18: Recommendations for Project Management

    • Propose recommendations for improving the project's intercultural communication and teamwork, based on your sessions 7 and 8 findings
    • Reference your brainstorming technique choices and how they integrate into the overall project management strategy

    Slide 19: Conclusion

    • Recap the main takeaways from your analysis
    • Reiterate the importance of understanding brainstorming techniques and cultural differences for successful global project management

    Slide 20: References

    • List sources and references used in the presentation

    Throughout the presentation, use bullet points, visuals, and concise text to convey the information effectively. Maintain a consistent design theme and font style for a professional appearance.

  7. User

    In the best possible way, give me a more detailed structure for my ppt presentation. Break it down into slides. The chosen global project is below, with all the information available.

    Project Title:AI IN ACTION (HOW THE HONGKONG DEVELOPMENT BUREAU BUIL THE PSS — AN EARLY - WARNING - SIGN SYSTEM FOR PUBLIC WORKS PROJECTS)

    EXECUTIVE SUMMARY Hong Kong has a large volume of construction works currently taking place; this will increase further and faster in the coming years. Hong Kong’s projects typically come in 15% under budget. However, ten per cent of local projects over-run their budgets, with a total overspend of eight percent. Some flagship projects have over-run their budgets and completion dates, causing public embarrassment. It is crucial to get these cost and schedule estimates under control in order that the booming infrastructure environment in Hong Kong can flourish to its potential while carefully monitoring costs and performance. We investigated whether Artificial Intelligence (AI) can play a part in accurately predicting these outcomes and identifying ahead of time when a project is going off track, allowing early intervention. The large volume of projects in Hong Kong needs to be monitored collectively, with a system that allows leaders to focus their efforts on the projects most likely to fail. An automated system simplifies this process, and in our studies was found to be as accurate as the best-in-class human cost estimates. High quality data is available in Hong Kong, and we studied a total of 2,700 years of combined construction activity and HKD 450bn in construction spent to trial this innovative project management approach. We unexpectedly discovered, however, that the AI needs only a very small amount of data in order to make good predictions. WE BELIEVE THESE EARLY FINDINGS WILL INFORM A NEW APPROACH TO PROJECT PERFORMANCE MONITORING AND LEAD THE WAY IN CAPITAL WORKS PROJECTS INTERNATIONALLY

    THE CHALLENGE Hong Kong expects project spending to increase by 42% between 2021 and 2031. The expected annual expenditure will reach up to HKD 325bn by 2031. However, these projects are not without their problems. To spot challenges early, the Project Strategy and Governance Office established a system to support supervision and governance of projects by providing early-warning signals to senior leaders of projects that are at risk. Projects are used to deliver the most significant initiatives in the world; from the Olympics, the artificial island of Lantau, to the International Space Station. One of the challenges currently faced by project leaders and investors is that projects do not perform well. They underperform in terms of execution, with significant cost and schedule overruns. 50% delays are common. Our research has shown this woeful underperformance is so consistent that we’ve coined the iron law of megaprojects: they are ‘over budget, over time, over and over again’. This poor performance is not a recent phenomenon: it has been fairly consistent for over 70 years. It is also common to different countries, sectors and delivery in the private and public sectors. The scale of projects has been increasing for decades. The largest megaprojects have budgets that match the gross domestic products of the largest countries. The GDP of Hong Kong in 2020 was USD 347bn, a similar amount to the budget for China’s high speed rail of USD 300bn and the USA’s joint strike fighter (USD 400bn). This large scale means that it is crucial for the money spent on projects to be better managed. The impacts of funds of this size being mismanaged can be felt on a national and even international scale. Over the period 2008 to 2017, performance on Government projects under the Capital Works Programme has generally been to a good standard, particularly from a budget perspective. Although certain projects required additional funding owing to project specific circumstances, the majority of approved projects were delivered at or under the original Approved Project Estimates (APE). Additional funding was required in approximately 10% of the projects and the amount represented some 8% of the total provision of these projects. Also of importance, this period of time was particularly active for the Industry, with approximately 850 Category A projects having their final accounts settled. In terms of cost, the original APE of these projects totalled approximately HKD 240bn as compared with cumulative final accounts of approximately HKD 210bn – leading to an overall surplus of HKD 30bn. Effectively, this means the surplus generated at the portfolio level was more than able to offset the cost overruns incurred by a select number of projects. To reinforce the positive nature of this performance, Professor Bent Flyvbjerg of the University of Oxford pointed out that Hong Kong is better than many other regions when it comes to project cost estimation, based on his findings in a study covering over 100 international jurisdictions. Megaprojects are those that come with high levels of procurement, design and/or construction complexity as well as scale. They typically take multiple years to implement and often involve high levels of risk to the participants involved. Unfortunately, a number of complex megaprojects in Hong Kong have suffered from underperformance in the form of delays, cost overruns and/or quality failures in recent years. These incidents have been well documented through extensive media coverage – leading to reduced levels of public confidence in the industry. The early stages of a project are crucial and can set the tone and influence its efficacy. One of the challenges of project performance is the sensitivity to timescales in taking corrective action. The later such actions are undertaken, the harder and more resource intensive it is to rectify problems. However, it is not always easy to identify risky projects before things have gone wrong and they require substantial and costly remedy. The Construction Industry Council estimates that construction expenditure will increase in real terms by up to 42% to HKD 325bn, compared to 2021. The Development Bureau of the Hong Kong Special Administrative Region has launched their vision for Construction 2.0 to support effective delivery. A key policy objective of the Development Bureau is ‘to ensure the effective planning, management and implementation of public sector infrastructure development and works programmes in a safe, timely and cost-effective manner and to maintain high quality and standards.’ In order to do this, it is vital for senior managers to notice problems and ensure efficient, effective and timely intervention when necessary. However, senior managers lack the daily and deep involvement in their projects. They need the right monitoring tools and early-warning -sign systems to help them in their project supervision.

    THE EXISTING SYSTEM EARLY WARNINGS A WEB-BASED MONITORING SYSTEM BASED ON CASHFLOWS OF HISTORIC PROJECTS The Development Bureau uses the PSS, or Project Supervision System, to monitor capital works projects and provide early-warning signals to senior management. The PSS is a web-based application comprising a series of standardized project cashflow graphs. The graphs are based on the past cashflows of capital works projects. The PSS covers different categories for building and non-building projects with various construction durations. All active projects submit their actual cashflows to date and their planned cashflows for the remainder of the project. The PSS uses this information to identify any unusual patterns. There are 3 types of patterns which then trigger different followup actions. First, normal performance. The project cashflow falls within the green zone. There is no follow-up action for projects with normal performance. The data of these projects will be stored for future use and refinement of the system. Second, yellow performance raises an alert. The project cashflow falls within the yellow zone, meaning heads of departments need to review these projects. The review takes in the risk of cost overrun and programme delay. The heads of departments are responsible for the project assessment and, if necessary, for the planning and implementation of appropriate recovery measures. Third, red performance flags the strongest warning. When the project cashflow falls within the red zone, the Permanent Secretary for Development (Works) chairs a top management meeting, which monitors all projects in this category.

    The monitoring system provides a better understanding of various elements of a project’s performance. Here are some insightful examples from the first year of using the system. The first project falls into the lower yellow zone, which indicates slow progress. The question to investigate is whether the project might be able to take actions to catch up (purple dotted line), or will be delayed (blue dotted line). The second project falls into the upper yellow zone, which indicates fast progress or cost overrun. The team needs to investigate whether the project is indeed making very good progress and will be completed early (purple dotted line), or if the project will experience cost overrun (blue dotted line). In either case project plans and budgets might need to be adjusted as a result of the investigation. The third project falls into the lower red zone and has stayed persistently in this zone for more than a year. This can signal a very long delay or an overinflated budget due to excessive contingency. In this case, the team investigated the project indepth and found that the project carried too much contingency. The project team reviewed the cost and developed a more realistic project budget. Once the cashflow was adjusted according to the new budget, the cashflow fell back into the normal zone.

    After using our monitoring system for over a year, we discovered that instead of it being a mere monitoring tool, the system provides a benchmark and even means to set better target for the project teams’ delivery of their projects. The data of these projects, once completed, will form part of the database creating a self-learning and self-updating system. Hong Kong’s project governance will be enhanced by the positive and interactive feedback between the PSS and the ongoing projects.

    THE DATA These closed projects have a total outturn cost of HKD 460bn (in 2019 HKD terms) and individual projects range from HKD 10m to 13bn. The mean project value is HKD 550m. The project durations range from 10 months to 18 years, with an average 3-year duration. In total, the portfolio contains more than 2,700 years of construction activity. The average project costs 15% less than forecast, and the range of cost outcomes is from 69% below the budget to 59% over budget. There’s a far greater range in project duration, ranging from 41% shorter than planned to 317% longer than planned. The average delay is 34% for the delayed projects. The historic 2008-17 dataset is representative of current performance, according to the Development Bureau’s response to the Legislative Council for projects in 2018-19.

    FIRST FINDING OF THE AI PROJECT TYPES AI ALGORITHM IDENTIFIED THE GROUPINGS OF PROJECTS THAT HAVE A DISTINCTIVE CASHFLOW PATTERN To analyse the cashflow data, two methods were selected: tracking annual performance as a percentage of the (1) outturn performance and (2) the original plan. A relative timeline was determined for each project, with the percentage of the total duration assigned to each year. For instance, in a four-year project the first year would have 25% progress, the second 50% etc. The data for the analysis comprised completed projects containing actual outcomes, but the forecasted outcomes will be used in the future, so the dual reporting will reflect cost and schedule performance as a percentage of the most recent forecast and the original plan. The analysis used random forrests, an artificial intelligence method to analyse the data. This unsupervised learning algorithm considered all possible characteristics to group the projects then extracted the typical cashflows from the s-curves and compared the fit of the curves. Analysis of the PSS identified that four types of projects had statistically significantly different outcomes. The analysis resulted in a decision tree, where the first decision divides all projects along the lines of building or non-building projects. Then the building projects are categorised by whether the original budget was smaller than HKD 150m or not and the non-building projects were categorised by whether the planned duration was less than six years or not.

    RESULTS FROM THE AI W A R N I N G FLAGS Each project was assigned a warning flag status; red, amber or green based on the outcome of the project. If the project finished in the middle 40% of the historic projects (30th–70th percentile) the project got a green flag. The bottom and top 5% of the project got a red flag (<5th or >95th percentile). All other projects got an amber warning flag (5th–30th and 70th–95th percentile). After assigning the warning flag based on the project outcomes we trained three AI algorithms (gradient booster, neural network, rolling discretizer) to find out the warning flag status based on progress and cashflows. The first AI assigns a warning flag for cost performance. The algorithm looks at progress and then the share of dispensed cash to date The second AI assigns a warning flag for schedule performance. The algorithm looks at the share of dispensed cash and then at the progress made The rolling discretizer, an algorithm specifically created for this task, achieved the most accurate predictions in unseen test data. To validate the predictive power of the algorithm the data were split into a training and a test set. The AI learned from 70% of the data how to assign warning flags. Then the accuracy of the warning flags was measured on the remaining 30% of the data, which the algorithm had not seen before. The AI achieves an excellent performance, especially in identifying the red projects correctly. The performance is well balanced across the different flags. 53%–69% of all the flags assigned to the observations are accurate. This means, for example, that if the AI assigns a red flag for a cost overrun the algorithm is accurate nearly 7 out of 10 times. The accuracy of the schedule flags is slightly worse. Here the AI is accurate in its prediction 18%–63% of the time. The AI outperforms the split by project type, which was only accurate in 8%–40% of predictions. The AI vastly outperforms randomness. If we simply made an uninformed “lucky” guess about the project’s risk we would get it right in 3%–25% of times. Especially for the very high risk category “red” the AI is more than 20–times better than a lucky guess. In addition to the validation with a test and training set of the data, we tested how little data the AI needs to achieve this level of accuracy. The analysis shows that is possible with as little as 10% of the data. We were surprised by this finding, because we always assumed that AI needs lots of data. It turns out we were wrong. Any organisation wishing to integrate AI into their early -warning-sign system needs some but not vast amounts of data. The AI is far from perfect; the accuracy is surpsingly good even though the algorithm only knew the cashflow and type of a project. Adding more data in the future will certainly improve the accuracy

    L O O K I N G F O R W A R D WITH THE AI P R E D I C T I N G COST The accuracy of the warning flag AI is promising enough to train more algorithms on an even more detailed task: forecast the outturn cost of projects. A more complex AI, a deep artificial neural network, predicted the final outturn cost with an average error of only ±8%. The accuracy is as good as a Class 1 estimate. The next challenge we set for the AI algorithms was to predict the final outturn cost of projects. The AI only knew the percentage of progress the project had achieved so far, measured against the planned project duration at final investment decision and the cashflow spent to date as percentage of the budget, that was approved at the final investment decision. For this challenge we tested many different AI and non-AI algorithms, including linear regressions, support vector machines, gradient boosters, random forrest regressions, k-nearest neighbours, multilayer percepetrons, simple deep neural networks and complex deep neural networks. We measured the error of the predicted outturn cost in the training data, which tells us how good the AI is at summarising the data it knows. We then gave the AI unseen data to test the prediction. As before, the AI used 70% of the data to train and 30% for testing. The best performing algorithm was the simple deep neural network. The mean squared error of the predictions for the test data was 1.6 percentage points. The mean absolute error in the test data was 7.9 percentage points. This means that on average the forecasted outturn cost was within ±8% at the portfolio level. The mean squared error, which is similar to the standard deviation, is only ±1.6%, thus 68% of predictions are within a range of ±1.6%, 95% of predictions are within a range of ±3.2% and 99.7% of predictions are within ±4.8% of the predicted cost. The mean absolute error of the prediction is ±7.9%, therefore an indication of the skew in the data, where very large overruns and fat tails are not uncommon. The AI estimate is in line with what conventional cost estimation practice considers to be the best cost estimate achievable. The Association for the Advancement of Cost Engineering (AACE) suggests that international best practice for a best-in-class cost estimate has an error between ±3% and ±15%; to achieve an estimate with this narrow error range, 100% of design needs to be completed and detailed information of unit cost and prices needs to be available. The estimate of the AI falls within the range expected for a Class 1 cost estimate.

    CASE STUDIES SPOTTING RISK ALL HONG KONG PROJECTS ARE MONITORED BY THE PSS. THESE ARE TWO REAL-LIFE CASES WHERE THE AI SPOTTED RISKY PROJECTS EARLY We selected two capital works projects from the portfolio to run the AI analyses. Project 1 has an approved project cost estimate (APE) of HKD 5.2bn with completion within 3.5 years. Project 2 has an approved budget of HKD 1.3bn with a construction duration of 3.9 years. Project 1 has completed its first year of construction. Project 2 has completed its first two years of construction. The latest cashflow information shows that project 1 has spent 1% of its cash and is nearly 29% complete when measured against the originally planned duration. Project 2 is 51% complete and has spent 21% of its cash. First, the AI assigned warning flags. Project 1 has a red flag. Project 1 has been spending the budget slower than the AI anticipates based on the typical spending pattern of capital works projects in Hong Kong. The project’s own forecast of spend (dotted line) shows that this trend will reverse in the next two years and exceed usual patterns in years 4 and 5. The AI’s flag raises questions about this unusual profile of acceleration and suggests further investigation. Project 2 receives a yellow flag. The project is spending its budget more slowly than typical projects but it forecasts that it will soon be in line with others. The schedule flags show a similar pattern. Project 1 gets a red flag for being very late if it can keep up the expenditure as is (under the assumption that this is not indicating a big cost underrun). Project 2 is progressing like a typical project in Hong Kong and the AI assigned a green schedule flag. But the AI raises questions about the plans after year 5. The AI also produces a cost forecast of the final outturn cost. At the mo- ment Project 1 is forecasted to cost HKD 4.0bn. For this estimate, the AI assumes that the project will not follow its own forecasted trajectory but behave more like a typical project in the future. However, the unusual pattern of expenditure means that if the project does indeed accelerate its spend the cost are likely HKD 5.1bn close to the APE. Project 2 has similar issues. The AI forecasts outturn cost of HKD 1.0bn. The project’s own forecasts show a spike in spending near the end; if that holds true the AI forecasts spending of HKD 1.3bn. Again the readiness for acceleration should be investigated.

    The enhancement of the PSS does not stop here. Our future plans include adding more detailed project data, further developing the AI to enhance the predictive power and extending the PSS to railway projects in collaboration with other international regions. Our innovations will not only support Hong Kong’s ambitious vision but also benefit projects worldwide.

  8. Assistant

    Given the provided global project information, here is a more detailed structure for your PowerPoint presentation, broken down into slides:

    Slide 1: Title Slide

    • Project title: AI in Action - Hong Kong Development Bureau's PSS
    • Team member names
    • Date

    Slide 2: Executive Summary

    • Importance of cost and schedule estimation for Hong Kong construction projects
    • Challenges faced by the industry
    • Objective of implementing AI to predict project outcomes and provide early-warning signals
    • Overview of findings and potential impact

    Slide 3: Project Background

    • Current situation and challenges in Hong Kong construction sector
    • Upcoming increase in project spending
    • Notable past project overruns

    Slide 4: Existing System (PSS)

    • Overview of the Project Supervision System
    • Purpose of PSS and how it is utilized for monitoring and providing early-warning signals
    • Explanation of the three types of performance (normal, yellow, and red)

    Slide 5: Using AI for Early-Warning Signals

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.