Construction Projects delivered On Time can no longer be SciFi
ROQUE.AI is disrupting the construction industry, empowering savings initiatives that could reach $1 trillion annually
For millennia construction projects run late
With ROQUE.AI, projects delivered On Time are no longer Science Fiction
Construction projects experience delays for millennia; be it the pyramids, the new Berlin airport or the HS2 in England, the outcome is always the same: projects are delivered late and at increased cost.
Multiple studies conclude that delays affect virtually all (70-90%) construction projects globally, with projects typically experiencing 20-30% time overruns on average, but with delays of up to 50% or longer being common in megaprojects.
Delays are also intrinsically linked to cost overruns, with basically all (80-98%) projects exceeding their budgets by an average of 20-30%, and megaprojects often seeing cost increases of 80% or more.
With the total size of the global construction activities currently exceeding $16.5 trillion and expected to surpass $20 trillion by 2029, the above overruns result also in trillions of money-waste.
We can do better! With the launching of the ROQUE.AI initiative we envisage to act as a catalyst for the transition to a construction world with more, less expensive, and faster delivered projects.
Time is Money
With very conservative calculations, halving of the experienced delays in construction projects could result in savings in excess of 6%.
For a $100 million project with a planned delivery within 24 months, this means potential savings of:
· $4 million on capex
· $2 million on opportunity waste / delay damages
· 4 months on human resources and machinery
Welcome to the ROQUE.AI SciRe(ality)
ROQUE.AI targets savings that could reach $1 trillion annually
Construction spending accounts for roughly 13% of the world’s GDP. Research shows that time and cost overruns have remained high and constant for the past 70 years. This persistence, despite decades of project management advances, suggests that the industry has been addressing symptoms rather than root causes.
Delays affect basically all construction projects globally (70-90%). The most frequent delay range has been measured to 21-30% of original delivery time, with significantly fewer projects (only 11%) experiencing delays of less than 10% — down from 28% in 2016. This trend indicates that delays are not only common but are becoming longer and more severe.
When projects extend beyond planned durations, they incur extended equipment rental costs, prolonged site supervision and administrative expenses, increased labour costs due to longer engagement periods, inflation impacts on materials and wages, potential penalties and liquidated damages, opportunity costs and financing charges etc.
With a very conservative calculation, halving of the above-mentioned delays in construction projects could result in savings in excess of 6% (see Analysis). For a $100 million project with a planned delivery within 24 months, this means potential savings of $4 million on capex, $2 million on opportunity waste / delay damages, and 4 months on human resources and machinery.
With resources being scarce and usually being blamed for late construction start, earlier completion of current projects could also bring forward future projects, potentially doubling annual industry growth, which currently stands at around 4.5%. This is more than 0.5% on global GDP!
This waste of money and opportunities should stop today! With the launching of the ROQUE.AI initiative, we envisage to bring together human experience and artificial intelligence to cut the gordian knot of construction overruns.
Beyond Money
Saving money is an important motive. And 0.5% additional growth globally every year could be a monumental achievement, unattainable by traditional policy making. But is it only that?
ROQUE.AI envisages to serve a broader and more ambitious vision: empowering the scarcer construction resource around, the experienced Project Managers, to deliver productivity enhancement, waste elimination, prosperity, and resilience to the world. A better world for the coming generations.
Our Vision embedded in ROQUE.AI's DNA
An AI-powered Augmented Reality for the Project Management Community
ROQUE.AI has been founded by professionals with long experience in the construction industry and Project Management in particular.
The initial aim of the ROQUE.AI Founding Partners was to utilise currently available and upcoming AI tools to supplement and facilitate human experience, to catalyse transformation towards a world with more, less expensive, and faster delivered construction projects.
By itself this is an aim the significance of which cannot and should not be underestimated: conservatively, $1 trillion could be saved annually, 0.5% could be added to the global GDP growth every year, and important future projects could be brought forward.
However, ROQUE.AI envisages to serve a broader and more ambitious vision: prosperity, resilience, productivity enhancement and waste elimination driven by the scarcer construction resource: the experienced Project Managers.
ROQUE.AI Founding Partners believe that this is a vision that could be served only by the construction community as a whole. Therefore we invite all willing professionals to be part of this disruption.
Anybody with a genuine interest in the construction industry or frustrated by its current underperformance is welcome to join the ROQUE.AI community. Either as a ROQUE.AI Partner, i.e. someone tasked to eliminate construction overruns, or as a ROQUE.AI Affiliate, i.e. someone interested to support or sponsor the ROQUE.AI initiative.
Why is Scheduling inefficient?
Experienced Planners spend tones of manhours drafting detailed Construction Schedules and still the delays persist.
Detailed Schedules seem incomprehensible to the non-experts and they continuously fall out-of-date, revised less frequently than necessary. Because their continuous amendment is considered unfeasible.
All the above limit significantly the potential for the Construction Schedules to efficiently identify and communicate projected overruns and to facilitate rational decision-making on appropriate mitigation.
The "Old World" of traditional Scheduling
Why Construction Scheduling remains inefficient
For every construction project, a proper and detailed Construction Schedule is (or should be) developed, long before breaking ground.
Each Construction Schedule attempts to consolidate all available (and projected) temporal information (delivery dates for material, commencement, duration and interdependencies of different workflows, constraints, milestones etc.) relevant to project execution.
A Schedule is developed by specialised “Planners”, utilising custom software tools, which is a task absorbing tens of manhours, especially for larger projects. The Schedule need be amended each time a single of the hundreds of its inputs changes, which is basically on a weekly basis. This continuous amendment is considered, on most occasions, practically unfeasible, with the Construction Schedules falling for long periods out of synchronisation with reality and, therefore, with their projections becoming unreliable.
Moreover, a Construction Schedule is usually presented to the project stakeholders in the so-called “Gantt” format, which is basically more difficult for the non-experts to follow, the more detailed and diligent it is.
The combination of incomprehensibility with the Schedule not being out-of-date on most occasions, limits significantly its potential to efficiently identify and communicate projected overruns and facilitate rational decision-making on appropriate mitigation.
Are delays inevitable?
According to the Normal Accident Theory, in systems which are characterised by high complexity and tight coupling, major failures are highly likely rather than unexpected.
Complexity is an intrinsic characteristic of construction projects, especially the larger they are.
But tight coupling is something we might be able to control or at least monitor closely. This is the focal point for the ROQUE.AI initiative.
Construction delays as a "Normal Accident"
Why construction projects are always significantly delayed
Time Overruns are driven by a large number of factors (see Report). The most important are:
- complex interdependencies between material deliveries and sequential subcontractors’ activities; even a single delivery delay or a temporary shortage of manpower or machinery or a short period of extreme weather could severely derail the overall project timeline, through a cascade effect.
- limited visibility on actual project status and progress; different stakeholders are rarely up-to-date or on the same page in relation to project status, given their “distance” to the site and, on most occasions, their non-technical background.
According to the Normal Accident Theory of Perrow (1925-2019), in systems which are characterised by high complexity and tight coupling, major failures are highly likely rather than unexpected. Construction projects share these two characteristics, hence the overruns experienced for millennia in most of them.
Complexity is an intrinsic characteristic of construction projects, especially the larger they are; but tight coupling is something we might be able to control or at least monitor closely. This is the focal point for the ROQUE.AI initiative.
The "New World" of ROQUE.AI
There are several measures which could be applied to control time overruns, but are usually considered unfeasible, due to high computational complexity.
But in the world of AI this might no longer be the case; utilising existing (and upcoming) state-of-the-art AI tools to supplement human experience, we could materially shrink computational and communication time, with reasonable cost, transforming On Time delivery of projects from SciFi to to SciRe.
AI supplementing Human Expertise
What can be done but is assumed currently practically impossible
cccectEIA.EtThe graph above illustrates the n: following an initial assessment, before breaking ground, a recursive feedback loop is established, upon commencement of the construction activities, which ensures early identification of delays and continuous, efficient mitigation, minimising time overruns.
· Feasibility analysis of the Construction Schedule: analysis e.g., if planned sequences are optimised, if spatial and temporal constraints on site are respected, if planned resources with their custom productivity rates match works’ duration shown etc.
· Sensitivity analysis of the Construction Schedule to circumstances which are not unlikely: assessment e.g. if contingencies and flexibility are incorporated to mitigate the effect of potential delayed deliveries of some long-lead items, extreme weather conditions, temporary shortage of staff or machinery etc.
· Technoeconomic analysis of potential mitigation measures: calculation e.g. if (and how much) additional manpower or machinery or overtime work could mitigate a week’s bad weather or if resequencing or spatial rearrangement could better mitigate the effect of a month’s delay in the delivery of a long lead item etc., facilitating rational decision making.
· Remote monitoring and visualisation of work’s progress: real-time automatic detection of construction delays, for early identification of potential overruns, and visual mapping of project progress, for efficient communication between all involved stakeholders and decision makers.
The accomplishment of the above tasks, within reasonable timeframe and cost, is no longer “impossible”; it could be achieved with the utilisation of existing (and upcoming) state-of-the-art AI tools to supplement human experience and materially shrink computational and communication time.
This is the foundational principle for the ROQUE.AI initiative (human experience and artificial intelligence working hand-in-hand, complementing one another), with its interlinked Diagnosis, Cure, and VisualTwin concepts.
The graph above illustrates the ROQUE.AI concept: following an initial assessment, before breaking ground, a recursive feedback loop is established, upon commencement of the construction activities, which ensures early identification of delays and continuous, efficient mitigation, minimising time overruns.
Optioneering on Schedules
As we no longer do multiplications by hand but rather use our calculators, we could also optioneer on Construction Schedules utilising state-of-the-art AI algorithms, a process, which in ROQUE.AI, we call Construction Schedule Simulation (CSS).
There already exist powerful AI tools (and further will be coming), which could be utilised for a CCS, and the ROQUE.AI initiative aims to support experienced Project Managers on the selection of, training on and efficient use of the proper armoury for every type of construction project.
ROQUE.AI Diagnosis & Cure
Construction Schedule Simulations and Resource Library
Each Construction Schedule amendment necessitates the engagement of an experienced Planner and the consumption of tens of manhours, especially for larger projects. This is the reason why Schedules are rarely:
· optimised, as investigation of many different possible scenarios (of workflows’ sequencing, resources allocation, tempo-spatial constraints etc.), the process we call “optioneering”, would necessitate spending of time and effort, which are usually unavailable or considered “unreasonable”.
· synchronised to current project reality, as their continuous amendment to reflect the current project status and up-to-date information on Schedule’s inputs (delivery dates, productivity rates etc.), would necessitate even larger effort to the one described above, for similar reasons.
But we can currently do better! As we no longer do multiplications by hand but rather use our calculators, we could also optioneer on Construction Schedules utilising state-of-the-art AI algorithms, a process, which in ROQUE.AI, we call Construction Schedule Simulation (CSS).
There already exist powerful AI tools (and further will be coming), which could be utilised for a CCS, and the ROQUE.AI initiative aims to support experienced Project Managers (PMs) on the selection of, training on and efficient use of the proper armoury for every type of construction project.
With the suitable CSS a PM could simulate deterministically multiple “what if” scenarios and calculate cost and time impact of any changes in a Schedule triggered by non-diligent planning, unforeseen events or potential mitigation measures.
However, the CSS could only work efficiently if it is driven utilising the ROQUE.AI ResourceLibrary, an extended proprietary construction database of the ROQUE.AI initiative, on necessary recipes, trades, resources and their productivity rates and costs, per work and project type.
In each iteration, a CSS calculates the optimum Schedule, which respects set project priorities and assumed resources, productivity rates, interdependencies and constraints. The whole optioneering process can be completed in just a fraction of the time and cost necessary by traditional means, making the whole concept feasible.
The graph above illustrates examples of risk identification and mitigation in a project (the original Schedule, represented by the green dot, is used as a reference point in this Duration vs Cost plot):
Pre-construction:
- ROQUE.AI Diagnosis: Feasibility analysis on the original Schedule identified inadequacy of planned resources for the timely completion of some works, based on productivity rates of the ROQUE.AI ResourceLibrary. The expected cost and time overruns, as a result of a cascading effect on the Schedule, were then calculated (represented by the blue dot), with the use of a CSS.
- ROQUE.AI Cure: The exactly adequate increase of resources to meet the original deadline was calculated (represented by the light blue dot), through iterative CSSs, utilising also input form the ROQUE.AI ResourceLibrary.
During construction:
- ROQUE.AI Diagnosis: The effect of an actual delivery delay on the Schedule was calculated (represented by the red dot), with the use of a CSS.
- ROQUE.AI Cure: Successive iterative CSSs, utilising input from the ROQUE.AI ResourceLibrary and different blends of mitigation measures, calculated the cost and time impact of scenarios serving different project priorities:
> The yellow dot represents the exact decrease of resources to match the original budget, in a cost-sensitive project.
> The grey dot represents the exact mitigation blend to meet the original deadline, in a time-sensitive project.
> The green dot represents the blend for an optimised cost-duration approach.
> In all cases, “overcompensation” scenarios (an example represented by the brown dot) were avoided.
Be friends with the "Big Brother"
As we can now pass passport control through automated unmanned gates, we could also utilise state-of-the-art AI image recognition algorithms to identify useful information on imagery of any type, a process, which in ROQUE.AI, we call Remote Visual Analysis (RVA).
There already exist powerful AI tools (and further will be coming), which could be utilised for a RVA, and the ROQUE.AI initiative aims to support experienced Project Managers on the selection of, training on and efficient use of the proper armoury for every type of construction project.
ROQUE.AI VisualTwin
Remote Visual Analyses and VisualTwin
Monitoring of actual project progress during construction involves regular reporting form site teams. It is a process that is time consuming and subject to human error; it not only lacks accuracy but also rarely brings any real value to project stakeholders, as it is usually based on exchange of long spreadsheets with numerous data and random photos, incomprehensible by someone without a technical background. Above all, collected data serve limited scope in predicting future status, due to the intrinsic complexity of construction progress: a delay in a workflow, even if identified, could have anything from zero to catastrophic effect on project completion.
But we can currently do better! As we can now pass passport control through automated unmanned gates, we could also utilise state-of-the-art AI image recognition algorithms to identify useful information on imagery of any type, a process, which in ROQUE.AI, we call Remote Visual Analysis (RVA).
There already exist powerful AI tools (and further will be coming), which could be utilised for a RVA, and the ROQUE.AI initiative aims to support experienced Project Managers (PMs) on the selection of, training on and efficient use of the proper armoury for every type of construction project.
A RVA tool utilises state-of-the-art AI image recognition algorithms, trained on actual project imagery, to process satellite or drone footage, for open-space projects (e.g. renewables), or 360 deg photo or video footage, for inner-space projects (e.g. buildings), in order to provide reliable, up-to-date remote monitoring and analysis of project progress status, allowing also for automatic detection of time and quality deviations during construction.
For open-space projects, satellite imagery could be readily available upon request, eliminating all complications related to permitting, safety, pilot etc. requirements of drone imagery. A RVA on satellite imagery could provide a clear overview of project progress, in material terms. When long periods of bad weather are experienced, drone imagery could be used in replacement. When greater detail is necessary, for technical reasons, drone imagery at high frequency, along with 3D modelling, could be utilised, to track construction activities with cm precision.
For inner-space projects, consistent locations of the building could be scanned, by human or robotic means, at a regular cadence, and the acquired imagery be put within the proper tempo-spatial context, archived, and referenced to the project’s detail design (be it in BIM or simple 2D format). This could allow, among others, "back-in-time" travel with a reliable historical record (no need to guess or “open up” a wall to see what is installed behind it), "before-and-after" visualisations (same views on different dates), progress reporting (areas that are on track, delayed, or ahead of schedule), or even quality monitoring (deviations to expected form, position, quality or quantity).
A RVA output, might improve our visibility on project status, but its value could be tremendously enhanced, if it is fed into, and consolidated within, regardless of its origin (be it satellite, drone, ground level footage or mix), the unique progress tracking and visualisation platform ROQUE.AI VisualTwin, to check and visualise works’ progress and investigate and check details on pace across trades and tasks.
On the foreground, the ROQUE.AI VisualTwin platform provides simple (at-the-touch-of-a-button), regular (at the preferred frequency) and reliable (up-to-date) tempo-spatial visual representation of the actual current, past and projected future project status, comprehensible even by non-technical teams, establishing common grounds among all project stakeholders and enabling more agile and effective management.
In the background, the ROQUE.AI VisualTwin platform maintains extensive, tempo-spatially referenced, historic, consolidated data, valuable even for the operational phase, and integrates with the Construction Schedule and the ROQUE.AI Diagnosis and Cure processes to early identify potential overruns and define and evaluate optimum mitigation strategies, in an ongoing iterative loop, until the project is delivered (on time and with proper quality).
There is no excuse for reluctance
Each day we do not take action costs almost $3 billions to the world; there is no time to spare. Review our Additional Resources, but, above all, ACT!
Get in touch. Regardless if you are a project Owner or a Contractor, an experienced or a junior PM, a graduate or an executive, you all have something to benefit and contribute to the cause. We need ALL to disrupt the construction industry and change the world!
Additional Resources
