BKWAI Insights

Geospatial Analytics

BKWAI — Fri, 04 Aug 2023 00:00:00 GMT

Geospatial Analytics – from Retrospective to Predictive and Why You Should Care?

Asset and construction monitoring has evolved at an arguably slow pace over the past few centuries. While it is very rare today to see a level and staff being manually used to determine the height difference between two points, it is still common to see a Total Station being used for the same purpose – a machine that has been around since the early 1970s. Both of these devices rely on principles devised by the ancient Greeks – while the hardware has of course progressed, the basic calculations have not.

What has evolved significantly faster though is our ability to interpret results from these data-capture devices and more importantly, make better meaning of the data, converting it into knowledge that can then be used for purposes such as asset behaviour diagnostics.

In conjunction, today we live in a world where data capture isn’t our bottleneck anymore either – it’s the complete opposite – vast volumes of data are captured both manually and automatically that then require interpretation and analysis. In the past twenty years, computing power has helped with this challenge.

As a result of this, geospatial platforms are now moving away from simple displaying of raw data to advanced interpretation of data. However, this is still a stepping stone. Most analysis focuses on the retrospective i.e., interpreting asset behaviour in the past. But what about looking beyond the past? What if we could use historical behavioural patterns to better help predict the future state of an asset, modelling when an asset may fail or, better still, determining the most appropriate time for maintenance investment?

Retrospective Geospatial Analysis

One of the hardest challenges with analysing geospatial data is trying to determine what the data is really telling you. Looking at a time-series of say vertical movement of an asset that contains thousands and thousands of points over time requires human patience. Fortunately, this can be augmented by some basic computing power.

Algorithms exist today that can help quickly remove noise from large data sets. Figure 1 shows an example of a time-series movement chart for an asset over a 6-year period. As an analyst, I must spend time and effort trying to manually interpret this noisy and complex line.

Figure 1 – 6 Year Time Series Line

To help me with the interpretation of this large dataset, I can apply a smoothing algorithm that reduces the noise and drives insight towards bigger picture trends over time. This is extremely helpful as it then gives me insight into trend variation without being skewed by local noise or abnormalities.

In the example below, I can see there is a downward trend up until July 2018 and then a reverse upward trend in the vertical movement of the asset in question.

Figure 2 – Trend-Change Determination

Such trend determination is helpful when considering drivers such as climate change and other long-term influencers but what about insight into shorter-term influencers? This becomes even more important during the construction phase of an asset where there can be influence changes on a daily or even hourly basis.

In the same way we can determine long-term trend change, we can also use simple algorithms to determine short-term step changes.

Figure 3 – Short-Term Step-Change Detection

In Figure 3 above, the computer has identified two points in time where there are discrete changes in asset height. These changes are not necessarily associated with the bigger-picture trend change. The first is the result of an adjacent asset failure – a wall failure that had a material impact on the asset. The second is because of a different root cause – construction itself. By overlaying the construction schedule, the analyst is able to pinpoint piling activities that occurred at the exact same time as the asset shift. From this, it can be determined, with a high degree of certainty, that the construction piling activities were the cause of the shift in asset height in early 2020.

Such root cause determination is possible due to the overlay of different dimensions – in this example, the asset movement and a construction schedule. However, the actual determination of the root cause (or correlation between the piling event and the asset movement event) is still largely down to the analyst’s expertise but even that responsibility is changing…

Today, mathematical correlation is being used to automatically determine relationships between assets and their environmental drivers such as construction activities. Analysts are being handed more and more guidance giving them more time to focus their expertise on making smart decisions rather than spending their time simply interpreting the data.

The real value of these advancements, however, is in the determination of the future state. We are on the cusp of being able to truly predict the future…

Predictive Geospatial Analysis

With the help of historical behavioural smoothing, trend detection, and discrete point-in-time change detection, we are becoming more proficient in being able to more accurately forecast future states through the use of extrapolation techniques.

There are numerous extrapolation techniques both algorithmic and statistical, and few of these are new. Table 1 below shows some examples of these.

Table 1 – Algorithmic Forecasting Techniques
Technique	Good for	Approach	Required data
Straight line	Non-noisy lines	Basic mathematics	Historical
Moving average	Cyclic lines	Basic mathematics	Historical
Simple linear regression	Primary driver e.g. temperature	Statistical	Relationship/observations
Multiple linear regression	Multiple drivers e.g. weather, traffic, ageing asset	Statistical	Relationship/observations

What is key though is that none is necessarily better than the others. At the end of the day, a forecast is just a model of a future state. If our asset movement over time is linear, then a straight-line extrapolation technique is appropriate; if the asset movement cycles over time, then it is better to adopt more of a moving average type approach. If there are multiple influences that are known to have a material impact on the asset, then we should consider more of a multiple linear regression approach. We need to apply an appropriate forecasting technique to fit the historical behaviour of the asset!

There is not a shortage of mathematically or statistically based approaches available at our fingertips. Again, that’s not our real challenge. Our real challenge is to figure out which forecasting technique is most appropriate for the asset in question. To best achieve this, we need to firstly revisit the historical behaviour and better understand the patterns to date to help us make better forecasting choices. This is where machine learning and AI can have a real impact.

BKWAI’s data science team have been developing technology to automate this forecasting selection process. As such, the ability to forecast more accurately is now a reality.

By looking across multiple assets, the most frequent root cause of historical behaviour can start to be determined. In other words, the drivers and root causes of the long-term trends and short-term step-change type behaviour are being better understood. With that understanding, software can be trained to adopt not just a specific forecasting technique but a relevant, and often different forecasting technique, depending on the asset, its age, its location and its maintenance strategy. This is a huge step forward in the science of forecasting and one that is already proving very useful. Asset managers are already benefiting from this improved forecasting to help with the maintenance strategy plans.

Figure 4 – Forecasting Example

As has been discussed, our ability to better interpret historical asset behaviour has been hugely augmented in recent years with advancements in geospatial analytical platforms. Going well beyond the basic rendering of raw data sets, we are now able to leverage interpretation techniques such as smoothing and even event-based changes at a given point in time.

The real benefit though of these advancements is the enablement of better forecasting. Forecasts can now be generated in the context of historical patterns, not only from the asset in question itself but from other analogous assets as well.

The next step in this evolution of analysis will be to capture the actual performance of assets as they progress through their lifecycle and use this to validate, and more importantly, calibrate as needed, our forecasting techniques for future analysis and monitoring. This is a far cry from, but still keeps true to, the centuries-old principles and usage of a level and staff! If we can better track and interpret asset behaviour, we have a much better chance of building and maintaining stronger, more efficient, and safer infrastructure.

Asset Management Support

BKWAI — Wed, 23 Nov 2022 00:00:00 GMT

Using technology to support Infrastructure Asset Management strategies

Written by Dan Patterson, Product Strategy at BKWAI

The Situation

Infrastructure Asset maintenance is a key part of the bigger asset management discipline. Determining appropriate response(s) to asset degradation helps to uphold the financial and service value of a physical asset.

An asset undergoes a life cycle that typically is represented through 4 stages – planning, acquisition, use and disposal. Each of these phases carries a cost as the example below shows.

There are well-established disciplines to help forecast and manage these cost phases including project planning for “plan”, construction management for “acquire”, maintenance management for “use” and decommissioning for “dispose”. Solutions like those from BKWAI fall into the discipline of maintenance management to support the “use” phase of the asset lifecycle. The use phase is typically not only the longest phase but also incurs the most cost as well. Understanding these two variables is key to successful asset management.

The Inevitable Asset Deterioration Curve

All assets deteriorate over time. The rate of deterioration is dependent on many influencers including:

Design characteristics of the asset
Environmental influencers e.g. normal weather patterns, soil conditions
Discrete events e.g. earthquakes
Activity e.g. traffic patterns

These all impact the speed at which an asset deteriorates.

Likewise, the rate of asset deterioration can be slowed down by maintenance intervention – this requires investment. Understanding how much investment is needed and when such investment should be applied is a difficult determination for asset owners to make given the uncertainty pertaining to the true future impact of the asset influencers. In summary, how do we achieve an appropriate mix of interventions in the context of influencers?

Types of Intervention/Maintenance

Maintenance strategies can be classified into three categories:

Preventative – proactive investment in scheduled maintenance activities resulting in reduction or even avoidance of degradation
Corrective – repairs to defects of minor impact after they occur
Reactive – response to a major defect after the fact

Corrective maintenance typically carries a 5x cost multiplier compared to an ongoing preventative maintenance schedule. Worse: a reactive response can be as much as 25 times that of a preventative maintenance plan. So why would an organisation ever not pursue a preventative maintenance plan? The answer boils down to the balance of risk-reward. The greater the forecast time window, the harder it is to predict an asset’s behaviour and how it will actually respond to its influencers e.g. 5 years from now, what will be the vertical movement on a pier? This makes the true determination of risk very hard to accurately predict.

If asset owners had greater visibility into the future state of their investments as well as a greater degree of confidence that the predicted state is indeed accurate, then the challenge of determining how much to spend on preventative maintenance becomes much more straightforward.

Using Data Science to supplement decision making

Using a predictive engine powered by data science algorithms allows us to forecast a future state by analysing historical trends and extrapolating these forwards to generate a predicted state at a given point in time. Achieving an accurate forecast can be complex depending on the number and magnitude of external influencers e.g. weather.

The more influencer dimensions available to the engine, the higher the degree of certainty that can be assumed from the model. Today, amongst other factors, weather and temperature influences help to better normalise forecasts.

By understanding the impact of preventative maintenance and comparing this to a strategy that rests its outcome on hoping corrective action isn’t required, we can easily calculate the ROI of these two strategies. From this, an informed decision can be made as to which strategy is most cost-effective.

Consider two simple scenarios: the preventative maintenance strategy and the corrective maintenance strategy.

In the example above, both strategies give us a 6-year asset life. So which strategy should we adopt?

Example Simple Cost Comparison Between Corrective and Preventative Maintenance
Strategy	Yearly preventative cost	Corrective costs	Total cost
Corrective strategy	$0	2 × $1MM	$2MM
Preventative strategy	$200,000	End of year 6	$1.2MM

Both strategies result in the same lifespan but the preventative strategy is $800K cheaper and results in better cashflow after year 3.

These simple models can be even more valuable when we add the dimension of certainty (or conversely, risk) to the equation. By modelling certainty of factors such as spend, effectiveness of corrective and preventative work on the performance of the asset, we can then start to report in the context of certainty as well. For example, a typical organisation will make decisions within the context of a given confidence or P value. Outcomes from different maintenance strategies will vary depending on the P-value adopted.

In the example above, a risk-averse organisation (one that adopts a high-certainty outcome e.g. P75) is better off adopting a preventative maintenance strategy (from a cost perspective).

Conversely, a more risk-tolerant organisation (willing to adopt a lower confidence of success e.g. P25) is better off adopting the corrective strategy.

Having such insight into alternate scenarios and outcomes is something that is very difficult to effectively achieve on paper or even in XLS-based models.

This is where products like those from BKWAI enable the merging of historical trend interrogation and predictive analytics to deliver powerful decision support to asset owners and operators.

To find out how BKWAI can enable you to make smarter decisions please contact us

Climate Change

BKWAI — Thu, 27 Oct 2022 00:00:00 GMT

Mitigating the risks that climate change poses for existing infrastructure

Written by Marisa Figueiredo, Head of Data Science at BKWAI

As the last few months of weather have adequately demonstrated, our weather is changing and is becoming ever more extreme. It’s likely that our mild winters and moderate summers will become the exception rather than the rule.

What does that mean for a bridge that might have an expected life of 120 years? Well, in all likelihood it was designed with the expectation that the climate would remain roughly comparable year on year, and while new structures can be designed and built with the knowledge that the climate is changing, the question that sounds alarm bells is, what will happen to existing structures and how do we protect them from damage and failure?

So, what are the risk factors?

In the UK specifically, although this is also true in some form for every country and continent, we are expected to see hotter, drier summers and wetter, warmer winters. In fact, according to the 2020 Global Climate Report from NOAA National Centers for Environmental Information, “every month of 2020 except December was in the top four warmest on record for that month”.

Not only is the trend line changing, but the frequency of extreme weather events is also increasing. Higher temperature spikes and more unpredictable and intense rainfall are of particular significance.

And what does that mean for infrastructure assets?

The impacts of these sorts of climate change phenomena are too numerous to mention them all, but examples could be:

Rainfall: More intense and frequent rainfall can change river flow conditions, increasing the risk of scour damage to bridge piers – scour damage is one of the primary causes of bridge failure worldwide.
Slippage: 2017 UK Climate Change Risk Assessment identified “Risks to transport networks from slope and embankment failure” (due to heavier and prolonged rainfall) as needing urgent action, specifically “to locate and remediate embankments and cuttings at risk of failure”. “Approximately 8% of the UK’s transport and road network is at medium to high risk of landslide disruption.” (Dawson, R.J., Thompson, D., Johns, D., Gosling, S., Chapman, L., Darch, G., Watson, G., Powrie, W., Bell, S., Paulson, K., Hughes, P., and Wood, R. (2016) UK Climate Change Risk Assessment Evidence Report: Chapter 4, Infrastructure.)
High winds and high temperatures: Can affect bridge stability, especially in historic structures.
Temperature spikes: Can result in operational downtime of rail networks due to tracks overheating, even creating a risk of rail buckling.

Consequences for asset owners

For asset owners and operators, the impact of climate on structures going unnoticed can vary hugely. In simple terms the earlier you know you have an issue, the easier and cheaper it is to alleviate it.

Once issues have had time to develop, assets often need to be closed or restricted for further inspections or in extreme circumstances, they can suffer partial or complete structural failure.

The implications of these can be wide-ranging.

Asset downtime and service disruption
Increased repair costs
Public and employee safety
Increased insurance premiums
Litigation
Reputational damage

Letting technology draw insights from the data

There are many ways to help alleviate these issues, from increasing the frequency and depth of manual inspections to closer monitoring of sensor data. However, these are expensive and time-consuming.

Platforms such as that from BKWAI allow the layering of multi-sourced data to provide a richer dataset from which the sophisticated algorithms can easily identify insights that would have taken hours, if not days, with manual analysis of the data. Our software delivers those insights straight into the engineers’ hands to enable them to make smarter decisions about the structure and its maintenance.

Imagine the situation: you operate an asset that has been in existence for 50 years. Six months ago, sensors were installed to check the movement of the bridge. Those sensors have identified movement. The only thing you now know is that there is movement; you do not have sufficient data to make an educated decision.

Now combine in-situ and satellite data, adding five years of historical InSAR measurements to the recent sensor readings. Now you have a layer of rich data that offers insights not available to you by any other means. Added to that, by layering additional dimensions such as weather, tidal movements, traffic patterns or even construction site activities, insight into root cause becomes a reality.

Comparing sensor and satellite data with the historical baseline helps separate true problems from normal behaviour. New or accelerating movement warrants investigation; recurring movement still needs assessment against engineering limits and the asset’s condition. This evidence informs inspections and preventative action.

In summary, modern I&M hardware combined with cutting-edge data analytics gives us more real-time access to asset movement, helps with calibrating what is ‘normal and expected behaviour’ and helps with determining root cause.

To find out how BKWAI can enable you to make smarter construction and asset monitoring decisions please contact us

Site Monitoring

BKWAI — Mon, 03 Oct 2022 00:00:00 GMT

Why Understanding Cause is More Important than Effect

Written by Andrew Covey, Head of Marketing at BKWAI

According to the June 2022 Geotechnical Instrumentation and Monitoring Market report, the 2022 global spend on geotech instrumentation and monitoring will top US$3B. In the UK alone, this equates to an expected spend close to US$100MM. Not only is this a huge financial investment, it is also reflective of just how much site monitoring data is being generated. So what exactly are we going to do with all this data and why is it so valuable to both contractors and asset owners?

The science of construction site monitoring is far from a new discipline – it has been around for centuries. However, the means by which we are able to capture site information, and the real-time nature in which it is captured and transmitted have dramatically changed. Automation has driven the cadence and quantity of data we are able to glean.

Having said that, one fundamental concept has remained constant. The primary data we are capturing today still pertains to the location of an entity – a location on a bridge; the corner of a building, the height of a railway track etc. On a construction site, we measure location through either a relative or absolute value and there are many standards in place such as the UK’s Ordnance Survey’s Benchmark system (height tracking) for tracking the movement of physical assets.

Yet with such an abundance of data comes a challenge – more data doesn’t automatically result in better insight or greater knowledge. There is a transformation that has to be made between data and knowledge. Data is a set of numbers; conversely, knowledge brings an understanding of what those numbers are telling us. While the I&M industry is, without question, experiencing a huge adoption boom, it is still in its infancy with regard to truly leveraging data to help better drive smarter decision-support. That is where site monitoring analytics come into play.

Site Monitoring Analytics

Site monitoring analytics can be separated into two parts:

Movement detection (the effect)
Root cause

Movement Detection

Firstly, by comparing positional data over time, we can determine not only movement but also the rate of change of such movement e.g. this asset is moving at a rate of 10mm per week. Again though, in order to be truly useful this insight needs to be presented with context – firstly, what is normal behaviour, and secondly, is this normal behaviour within our thresholds or design tolerances?

In order to establish what is normal behaviour, we need a benchmark and the best way of establishing such a benchmark is to look back through time and try and determine a historical trend.

By establishing such a benchmark, we can then compare recent movement against the trend so as to determine whether the change is out of ‘tolerance from the norm’ and, if so, whether it was caused by change over time or a discrete point-in-time event such as a vehicle collision.

With the advent of cloud computing and data mining, capturing and analysing very large quantities of data to help establish benchmarks is becoming an affordable reality. With the availability of such benchmarks, we can now make much better judgement calls as to whether a positional change should be of concern or not.

In the case where it is deemed to be a material concern, the real question to be answered is ‘what is the root cause of this change?’ If this can be answered, then we can develop a response to the driver of the change, hence de-risking the probability of a failure or damage. The value here is all about de-risking…

Determining the Root Cause of Movement

Engineers working on the construction of a super sewer in an area prone to flooding are using the BKWAI platform to help forecast flooding.

A change in position of an asset can be caused by or influenced by numerous factors (and worse, potentially compounded by a combination of factors). Daily cycles in temperature, traffic patterns, scheduled site work and an unforeseen incident are all examples of causes of movement.

The more context we can apply to a given entity, the better chance we have of pinpointing the root cause of its movement. This is where the concept of multi-dimensional analysis comes into play.

If we take a given sensor and overlay additional dimensions such as temperature, then we can start to determine whether the movement change is correlated with a corresponding change in the weather. This is relatively easy to do in the age of easy access to weather-related data.

Taking this a step further, we can overlay disparate data sources and tie them back to the time-phased data stream we get from a sensor, broadening our ability to truly determine root cause. A real-world example of this can be seen in the BKWAI view below that shows site activities being pulled from the construction schedule and overlaid onto the site movement chart. From this we can see that there is a direct link between pile driving activities and the asset experiencing a material and threshold-breaching change in position.

Pinpointing the root cause back to the pile driving activities not only enables us to record the event in the form of an auditable digital thread but also gives us the opportunity to remediate a subsequent analogous event in the future.

The concept of multi-dimensional root cause analysis is a major step forward in asset monitoring. For the first time, stakeholders can now get not only a single source of truth pertaining to an asset, but more importantly, get guidance on the cause of change to that asset. Weather and site activities were mentioned as examples above. Imagine the possibilities over time as we continue to digitise construction environments.

To find out how BKWAI can enable you to make smarter construction and asset monitoring decisions please contact us

Seeing The Whole Picture!

BKWAI — Wed, 03 Aug 2022 00:00:00 GMT

Looking at one aspect of a building or asset tells you only one thing; aligning and correlating multimodal data allows you to see much more.

Written by Andy Covey, Head of Marketing at BKWAI

Imagine you own a bridge, and your physical sensors tell you that there is movement. What do you do? You could close the bridge to investigate the movement, restrict the access while you carry out more monitoring or step back and look at a wider picture of the situation by combining in-situ and satellite data, including historical InSAR records.

Why does historical data make such a difference?

Different sensors tell you different things. Installing sensors on site can give high-frequency readings every few hours, minutes, or seconds in very localised regions to tell you what is happening now. What they don’t provide is context.

If you were blindfolded and presented with each of the items below, what might you think each one was?

Now you stand back and remove your blindfold; the truth is revealed.

Historical records establish normal behaviour, helping distinguish anomalies from expected responses to temperature and water levels.

When that information is coupled to the current sensor readings it is far more likely to reveal the true picture or at least insights into that picture.

The ideal solution is to use different tools for different jobs. The BKWAI platform can use machine learning to distinguish actual issues from normal behaviour, taking temperature and other environmental influences into account.

Is historical data the only time I can reveal the elephant?

No, one of the key benefits of the BKWAI platform is that it can bring in information from any brand or type of sensor as well as many other sources such as weather, to provide a single source of truth. As the effects of climate change become more evident, the integration of temperature and flood maps into the platform will offer invaluable insights into the truth of your asset, its site and the surrounding area.

Rather than logging into a platform for each type of sensor or the environmental data and viewing it in isolation, you see all the information in relation to the rest. Instead of looking at each of the elephant’s legs individually, we reveal the entire animal in a single view.

To find out how BKWAI can help you see the full picture please contact us

Gain Competitive Advantage

BKWAI — Thu, 09 Jun 2022 00:00:00 GMT

How BKWAI’s data science and engineering expertise is helping contractors gain a competitive advantage.

Written by Marisa Figueiredo, Head of Data Science at BKWAI

There has never been more data available to the construction and infrastructure industry for monitoring the built environment.

Satellites, sensors, environmental monitors and more are generating an abundance of high-quality data with the potential to be used to make smarter, safer, more sustainable and more efficient decisions across all aspects of the construction process.

But without the technology to harness this data, little of it is translated into value on site during procurement and construction.

Without a way to speedily extract salient insights from the overwhelming amount of data available, decisions are made using the same, often manual processes that they always have been.

Neither productivity nor profitability can be maximised when this is the case.

Potential savings from data analytics and related technologies may amount to as much as 10 to 15% of total construction costs, according to a report by Deloitte.

At a time when the construction and infrastructure industry’s margins are eroded (to just 3% in European companies), data science’s ability to minimise costs while meeting timelines and quality and security requirements can make a significant impact on profits.

BKWAI’s solution

BKWAI’s proprietary technology uses machine learning and advanced analytics to empower engineers with intelligent insights and analysis from relevant and actionable data, be it from sensors, satellites, environmental monitors or elsewhere.

The platform brings together on-the-ground sensor and satellite data with environmental and other monitoring streams.

Engineers can use these insights, presented on a single portal, to rapidly form a detailed and accurate understanding of how site activities are impacting the condition of monitored structures and spot on-site issues before they arise.

The benefits are numerous:

It is easier for contractors to overcome safety concerns.
Staff time is used more efficiently as engineers no longer need to delve through thousands or tens of thousands of lines of data looking for patterns.
Fewer site inspections are needed.
Issues are spotted before they become a problem, allowing project plans to stay on track.
Projects are less likely to be delayed as a result of unexpected issues needing to be rectified.
The costs of schedule overruns are reduced.

How BKWAI is reducing costs for asset owners and construction companies

Predict risks and increase safety

Engineers working on the construction of a super sewer in an area prone to flooding are using the BKWAI platform to help forecast flooding.

Real-time weather forecasts helped predict, with far more accuracy, events such as localised storms that cause the site to flood and avoid the costs and dangers of sudden evacuations.

It is now possible to predict weir levels and issue early warnings when a critical water level will be reached. As a result, site teams can make informed decisions about accessing the site, rather than having to react swiftly to flooding.

Previously, workers and equipment had to be rapidly evacuated from a 41m deep shaft whenever the site flooded – a huge inefficiency and a safety risk.

Spot issues before they become a problem

Engineers working on the construction of a major shopping centre within a four-storey Grade II-listed facade are using the BKWAI platform to monitor movement of the protected assets.

Even the slightest movement is detected quickly, long before it would be spotted otherwise, helping protect the historic facade and surrounding buildings.

Disruption and delay to the normal operation of the construction site are avoided by reducing the need for unexpected surveys or urgent unplanned works.

The platform also visualises how the structures are changing over time, so engineers can predict how they will behave in the future.

As the BKWAI platform brings in sensor data from many brands and types of sensors, engineers can use just one software platform to view and compare data from multiple sources.

The platform also aggregates the data, providing visualisations engineers use to rapidly home in on the specific data sets they need to look at, rather than delve into endless spreadsheets.

As a result, highly trained and expensive engineers spend more time making data-based decisions and less time searching for issues.

This leads to reduced costs, greater efficiencies and increased worker safety.

Save on procurement and reduce carbon emissions

When the largest environmental cost of a structure is replacing it, extending its lifespan maximises the value of the initial investment and the carbon impact of its construction.

The BKWAI platform is being used to show how a bridge and the surrounding landscape have been moving over the past 5 years.

Engineers are able to ascertain that the movement seen in the structure does not need to be addressed as the same movement occurs every year.

This insight meant remedial work to keep the bridge operational could be minimised.

The asset owners saved tonnes of additional concrete and steel, hundreds of thousands of pounds of site expenses, and reduced their carbon footprint.

To find out how BKWAI can enable you to make smarter construction and asset monitoring decisions please contact us

Identifying Cost Savings

BKWAI — Tue, 24 May 2022 00:00:00 GMT

Cost-saving insights from the BKWAI platform that were invisible to the human eye

Written by Sonja Abhyankar, Head of Product at BKWAI

When a monitored asset such as a bridge or a protected structure has moved so much as to be visible to the human eye, the time for preventative action has often passed and costly remediative intervention will be needed. The bridge will be closed for repairs. The protected structure has to be repaired.

Despite this, most monitored assets are visually inspected according to a fixed timetable. Although this offers a bespoke approach, small changes and emerging outliers can be missed. And if a significant movement happens just after a recent inspection, it can be years before it is spotted.

As a result, movements that indicate an underlying problem are spotted late, or unnecessary interventions are made when normal behaviour is mistaken for an underlying problem. In both cases, excess and avoidable costs are incurred.

BKWAI’s solution

BKWAI’s platform combines in-situ and satellite data: sensor and survey measurements provide local detail, while current and historical InSAR records provide the wider geographical context. Five years of satellite history add depth to the assessment.

The platform brings together on-the-ground sensor and satellite data with environmental and other monitoring streams.

It is deployed in the cloud and automatically ingests data from all relevant sensors, regardless of type or manufacturer, allowing simultaneous visualisation in one place.

Even the slightest movement between inspections can be detected and potential problems prevented. This helps asset owners understand trends, correlations and the root cause of these movements, and whether they need to intervene.

This can generate significant cost savings. A Deloitte report found that analytics-enabled initiatives generate a three to seven per cent increase in ROI due to the extended useful life of the assets.

Cost savings through early intervention

Engineers working on the construction of a major shopping centre within a four-storey Grade II-listed facade are using the BKWAI platform to monitor movement of the protected assets.

BKWAI’s constant monitoring allows them to see how site activities are impacting the condition of monitored structures.

Instead of having to rely on manual surveys and graphs plotted by a human, engineers can access intelligent visualisations and data analysis that allow them to precisely interrogate what is happening on the site. The system highlights issues, creates alerts and notifies engineers via email or SMS, so they no longer need to spend days buried in Excel spreadsheets looking for concerning readings.

As a result, engineers can spot on-site issues before they arise and need to be repaired, as well as ensure planning compliance.

Cost savings by avoiding unnecessary work

BKWAI’s platform analyses how a bridge responds to temperature, environmental conditions, wind, loads and construction activity. Machine learning helps separate anomalies and areas of concern from normal behaviour.

Five years of records established the bridge’s normal response to temperature, helping engineers distinguish it from anomalies and areas of concern.

This removes the need to do costly remedial work, as well as the disruption that would be caused by closing the bridge.

Cost savings from alerts for movement

Engineers working on the construction of a major stadium development within a Grade II-listed location are using the BKWAI platform to monitor movement of the historic structures and ensure compliance with planning rules. Configurable alerts trigger email or SMS notifications when any set movement parameters are breached, long before they would be visible to the human eye.

Movement across the site can be viewed on 2D and 3D maps and using satellite imagery. Each individual sensor or array of sensors can be quickly and easily interrogated to view historical changes.

As a result, intervention can be made before the protected assets move so much as to cause damage, avoiding the cost and delays of repairs, as well as expensive regulatory fines.

To find out how BKWAI can enable you to make smarter construction and asset monitoring decisions please contact us

Cambridge Centre for Smart Infrastructure and Construction – Smart Infrastructure Blog

BKWAI — Tue, 03 May 2022 00:00:00 GMT

First Published by Cambridge Centre for Smart Infrastructure and Construction – Visit the CSIC website

Our increasingly data-centric world brings opportunities to create a more resilient, resource-efficient and sustainable built environment. But in order to realise the benefits of insights delivered from big data, there needs to be greater attention paid to curation and management. Dr Sakthy Selvakumaran, co-founder of tech start-up and CSIC spin-out BKWAI and an Engineering Fellow based at the Civil Engineering Department at the University of Cambridge, sets out what’s at stake.

Global infrastructure in many countries is ageing and performing beyond its design life and intended capacity. Faced with the climate emergency and extreme weather events, the built and natural environments are under increasing strain and resilient infrastructure is needed to provide services on which society depends. Government targets for net zero require the built environment sector to better understand the condition of existing assets and maintain them using minimal resources. The value of data has long been the focus of CSIC’s work which demonstrates the value of data as an engineering tool for better decisions to improve the performance of our infrastructure and embed low-carbon, low-waste outcomes into projects with the potential to transform industry practice.

Sources of data

Data has never been so abundant or potentially valuable. The proliferation of new sensors and satellite technology has the potential to fuel a revolution in the way the built environment can be monitored. Data can be used to make smarter, safer, more sustainable and more efficient decisions across all aspects of the process – from planning to construction to operation and maintenance. In the built environment there are many and varied sources of data available in ever-increasing volumes and accessible forms. Data sources include sensors attached to structures, and embedded sensing systems – fibre optics installed in new structures and piles – which can help engineers to understand capacity over an asset’s lifetime and beyond. Data can also come from remote sensing, including satellite acquisitions, drone surveys and laser scanning as well as automated sensing systems, such as mixed reality solutions for construction progress monitoring.

“Data housework may not be exciting, but it is essential to optimising value – not just in the present but the future. Huge amounts of data are currently generated during construction and the operational management of assets, but are typically looked at once to decide a threshold before being forgotten or archived never to be looked at again, often becoming obsolete in a form that can’t be easily accessed. Dr Sakthy Selvakumaran, co-founder BKWAI”

Even data not intended for the built environment provide additional information – social media, geo-tagged photos and GPS from mobile phones can tell us about people and dynamic behaviours. The cost of sensors is dropping and free, full and open data policy by the European Space Agency (ESA) provides access to optical, radar, and other forms of satellite data that can help us track the dynamic behaviour of assets and the environment. In addition, there has been tremendous progress in spatial, spectral, temporal and radiometric resolution of data gathered by satellites providing an increasing volume of data to study structures on a global scale.

Opportunity of data abundance

The more tools an engineer has in their toolkit the better – built environment monitoring and measuring requirements come in all shapes and sizes depending on what is being measured and why. A multimodal approach makes use of the diversity of sources of data and combined, provides a better overall picture of an asset and its environment. Insights can support more informed asset decision making, reduce risk and enable proactive maintenance to assets in changing conditions brought by climate change.

The challenge of optimising data abundance

Access to more data does not deliver benefits immediately. Simply collecting data will not yield value – engineers and asset owners must make sense out of sensor data and understand how it can be used to gain insights to make smarter, safer, better decisions.

Here lies the challenge. Manual analysis of data by human operators is not efficient, except for very limited tasks, because the quantity of data is so vast and the associated labour is significant. The only feasible approach to enable data processing is large-scale automation, fusing together remote sensing data analysis with domain understanding. This is particularly true when investigation is carried out on dynamic and extreme scenarios, such as monitoring the changing aspects of urban scenarios and the potential environmental risks and hazards to predict problems – which allows infrastructure owners to take action to improve the resilience of the assets they are responsible for. Combining machine learning and deep learning techniques with high performing computing enables the automatic processing of large amounts of data to assess infrastructure at scale and spot trends and changes in behaviours, such as ground displacement, across an area of interest.

Curating data for future use

While the possibilities of emerging digital tools and approaches including data science, machine learning and artificial intelligence are compelling, they will remain beyond reach unless our data are in order. Organisations collecting data can only extract full value from curated data – we must consider how we store, structure and clean data so that it is accessible beyond the lifetime of a single project and useful to future asset owners and operators. When paper drawings were an important point of reference the drawings were filed and archived to be retrieved later. Similarly, sensor data must be curated for future use as a resource to compare and contrast and inform modelling and design.

Data housework may not be exciting, but it is essential to optimising value – not just in the present but the future. Huge amounts of data are currently generated during construction and the operational management of assets, but are typically looked at once to decide a threshold before being forgotten or archived never to be looked at again, often becoming obsolete in a form that can’t be easily accessed.

Valuing and managing data as an asset in itself is crucial. These curated building blocks form the foundations for big-data analysis, models, analytics and data mining which can provide insights to inform better decision making to support and improve the built environment systems and services on which we all depend.

BKWAI receives venture funding to accelerate data-driven infrastructure and construction

BKWAI — Wed, 02 Mar 2022 00:00:00 GMT

BKWAI is proud to announce that we’ve received £2.2m in Seed funding in a round led by Octopus Ventures, one of Europe’s largest and most active early-stage investors, with participation from Deeptech Labs and Cambridge Enterprise (the investment arm of the University of Cambridge).

Global infrastructure in many countries is ageing and performing beyond its design life and intended capacity. It is an industry that has 2-8% profit margins and is going through a period of rapid digitisation in the UK. This has been given an initial push by government regulation such as the mandate for BIM Level 2, and accelerated by requirements imposed by the COVID-19 pandemic.

Therefore, the construction industry needs to better understand how existing assets are performing, so it can get more out of them using minimal resources. However, assessing the condition of assets can be difficult. There are logistical challenges – visually inspecting remote and hard-to-access assets can be expensive and hazardous. Even in new construction, it is important to ensure safe environments and understand the short- and long-term environmental impact of a site. COVID-19 further exacerbated these challenges as it meant the number of people who were allowed on site was limited. In the longer term, the industry is also under pressures associated with the climate crisis and extreme weather events, which add an additional strain to the condition and structural health of assets, making assessing their condition increasingly important.

The BKWAI software platform combines data from multiple sources (on-site sensors, environmental data and remote satellite data) giving a whole view of the asset, as well as its surrounding environment. The company deploys advanced data techniques to help engineers sift through huge volumes of data to find patterns, predict what is going to happen and identify early warning signs. The platform puts this insight into the site engineer’s hands quickly and in real time, enabling construction companies and asset owners to make more timely, more cost-effective and safer decisions.

Although only founded in 2019, BKWAI is already deployed on construction sites by some of the biggest names in UK construction and on existing major infrastructure assets. Current customers include Thames Tideway, Laing O’Rourke, Kier, WSP and Highways England. BKWAI’s founder, Sakthy Selvakumaran, has over a decade of international experience in design, construction and maintenance of infrastructure assets, and she combines this with a PhD and academic research in satellite monitoring and data analysis methods in the University of Cambridge Engineering Department. The company is also supported by Geovation, part of Ordnance Survey, UK.

Akriti Dokania, Investor at Octopus Ventures, comments, “Construction is one of the world economy’s largest sectors, worth £7 trillion. McKinsey estimates that if the productivity levels were to catch up with the progress made by other sectors over the past 20 years or with the total economy, it could increase the industry’s value added by £1.2 trillion a year.

We’re delighted to be backing such a formidable founder who is taking on a giant industry and we’re excited by BKWAI because it is giving the industry the tools it needs to meet its potential. Not only is it helping companies optimise their builds and assets, it is also moving the industry to more green, sustainable practices that have long-term impact.”

BKWAI’s founder, Sakthy Selvakumaran, says, “The business is growing rapidly and is taking on investment to scale operations and build out its roadmap. It is a really exciting time at BKWAI. We are seeing increased demand and as such are expanding our offering to meet the industry’s needs. The proliferation of new sensors and satellite technology has the potential to fuel a revolution in the way the built environment can be monitored. Data can be used to make smarter, safer, more economical decisions across all aspects of the process – from planning to construction to operation and maintenance.

However, in an industry rooted in spreadsheets and manual interventions, it becomes impossible to sift through such volumes and find key insights and values. One sensor or source of information does not fix all problems – the industry is crying out for better tools to help understand the complex challenges facing our built environment. That’s where we step in – we can give them the much needed data-driven engineering insights, to find the right solutions and take action.”

Smart cities in a post-COVID world

BKWAI — Wed, 18 Aug 2021 00:00:00 GMT

COVID-19 has been a paradigm shift. It has changed the way we view the world, the way we connect and interact, and the way we work.

In addition to its more direct impact on people, it has affected the systems and environment that support us all. From safe water from our taps, to how we heat and light our homes, to the bridges, tunnels and transport we use to get to work – our infrastructure provides the backbone of modern UK society. For many of us, this is something we take for granted under normal conditions; but as COVID-19 took its grip, engineers began to form part of the critical workforce, ensuring that these fundamental pieces of the built environment stay safe and properly maintained. Despite the physical restrictions, construction must continue, assets must be constantly checked, and critical infrastructure must be conserved.

The pre-existing challenges of continual inspection, monitoring and maintenance by sheer numbers of people working on site, allied to omnipresent financial constraints, have led to the inevitable conclusion that a smarter solution is required.

As the world looks to the future, there has been a resurgence in the belief that smart cities may offer a viable way forward.

A smart city is determined as an intelligent network comprising myriad sensors and devices, connected via the internet and IT infrastructure. Receiving, analysing and managing data in real time effectively enables citizens to interact with smart city ecosystems to facilitate improvements in areas such as traffic flow and even air quality. In theory, this evolution results in performance-influenced decision making.

A recent report from US-based ESI ThoughtLab entitled “Smart City Solutions for a Riskier World” suggested that, as a result of the pandemic, around two thirds of the 167 cities surveyed worldwide were reconsidering urban planning.

“Our discussions with global city leaders have made it clear that the pandemic has been a catalyst for remarkable technological, business and social change, which will continue even after the health crisis ends,” announced ESI ThoughtLab CEO Lou Celi. “The most successful cities will be digitally transformed, citizen-centric, and fully sustainable.”

One of the American cities leading the field with its investment in new technology is Philadelphia, which is using artificial intelligence to improve the quality of one of the major bugbears of its residents – its pavements. The AI-powered ‘pavement repair prioritisation tool’ collects roadway quality data in order to save money and reduce repair times on the city’s streets.

The project covered 1,200 miles of pavement in its first three months and led to the development of similar initiatives mapping manhole locations and street signs.

But not all smart city projects have had such a positive outcome.

Last year Sidewalk Labs, a smart city subsidiary company of Google, slammed the brakes on its ambitious plans to build a sustainable, data-driven utopia in a water-fronted district of Toronto, Canada.

Sidewalk Labs CEO Dan Doctoroff cited “unprecedented economic uncertainty” in the wake of the pandemic as the deciding factor in pulling the plug on the Quayside scheme.

Projected to cost over one billion dollars, the venture was launched in 2017 and counted Canadian prime minister Justin Trudeau among its supporters. An artist’s impression of the waterfront – seemingly visualised in the middle of a kite festival – led to the project being dubbed ‘Google Island’.

Boasting ‘raincoats’ for buildings, autonomous vehicles and even heated pavements, the plan nevertheless ran into opposition from locals, concerned about the lack of clarity over the use of sensitive data.

There were many critics who saw the idea as a dystopian Valhalla akin to that imagined by HG Wells in ‘The Time Machine’ or, for younger readers, the Capitol in ‘The Hunger Games’.

So, beware the merchant that tries to sell a seamless vision of smart cities. Instead, cities need to start by asking residents to identify their needs and then consider all possible options without assuming that the answer to every problem involves big data and new technology.

More specifically, we need better tools to support skilled labour in making important decisions. The days of optically scanning spreadsheets have gone, replaced by state-of-the-art data science that enables engineers to plan more effectively, spot potential issues and minimise project disruption.

By supporting engineers across the asset lifecycle, the toolkit requires the capability to access historical satellite data, thus unearthing previously unseen insights. Software solutions need to be smart, secure and super-fast, using the thousands of sensor readings on site to present engineers with a fuller picture.

The geoscientists in the team can react quickly to environmental measurements, as the project’s skilled workforce receives a wealth of other high-volume data streams at high sampling rates.

Data is important; but data isn’t going to drive the revolution – people are. Let’s use data, harvest it, access insights and present it with useful perspectives so that smart people can make smart decisions.

What does climate change mean for existing infrastructure assets?

BKWAI — Mon, 21 Jun 2021 00:00:00 GMT

Infrastructure assets are designed to operate over the long term. The vulnerability of these assets due to age and deterioration is now being further jeopardised by climate change, which has become a topic of debate and research among engineers and policy makers.

Bridges, with an average design life of around 120 years, are one of the most relevant elements when considering the effects of climate change. Typically, they have been designed on the assumption that the climate of the future is very similar to our current one.

We can of course start to include the effects of climate change as we plan and design new infrastructure, but what about existing infrastructure that has already been built?

Let’s start our brief globe-trotting trip by looking at the impact on America. According to the 2020 Global Climate Report from NOAA National Center for Environmental Information covering a state-wide review of the US, “every month of 2020 except December was in the top four warmest on record for that month.”

As well as its warmer climes, America is feeling the financial impact on its federal resources and taxes which, as a consequence, has left many of the country’s bridges in a precarious position. A report published in 2019 noted that the integrity of many US bridges relied on regular and efficient maintenance to keep the expansion joints functional in order “to avoid undesirable thermal demands that would potentially be magnified due to climate change”. The report’s findings stressed the need to develop new materials and technologies that would prove both cost-effective and sufficiently flexible to enhance the resilience of the transport infrastructure.

In Scandinavia, the degradation of its infrastructure has started to accelerate due to its shorter, milder winters and longer, hotter summers. Sweden’s projected higher temperatures, increased rainfall and humidity, allied with higher carbon concentrations in the atmosphere, have led to an increased risk in the deterioration of its bridges. A recent report on carbonation-induced corrosion predicted a rise of more than 400% by the year 2100, citing climate change as the principal cause and stating that similar trends could be expected regarding steel bridges.

As well as the problems of increasing temperatures and more unpredictable rainfall, into the mix comes the heightened intensity and frequency of extreme weather events. Flooding is of particular concern, with shorter return periods and simulations predicting that some bridges may be totally submerged in the not-too-distant future.

More intense and frequent rainfall can change river flow conditions and increase the risk of scour damage to bridge piers. One of the primary causes of bridge failure worldwide, scour damage occurs when fast-moving water removes sediment from around the bridge foundation, leaving behind scour holes which, in turn, can seriously compromise the bridge’s integrity. If too much sediment is washed away, the bridge may collapse.

Published in 2017, the UK Climate Change Risk Assessment identified “Risks to transport networks from slope and embankment failure” (due to heavier and prolonged rainfall) as needing urgent action, specifically “to locate and remediate embankments and cuttings at risk of failure”, and identified that “approximately 8% of the UK’s transport and road network is at medium to high risk of landslide disruption.”

Add to this the effect of high winds on bridge stability and high temperatures that can result in operational downtime due to rail tracks overheating and even buckling, and it becomes clear how significantly climate change can increase the risk of structural and operational failure of infrastructure assets.

The consequences for asset owners are wide-ranging, from complete or partial structural failure, the considerable cost of repairs, increased insurance premiums and operational failure, to asset downtime and, in the case of the rail network, service disruption.

It also raises concerns over health and safety issues and, in broader terms, erodes public trust and impacts on reputation.

The value of smarter asset monitoring to offer an effective strategy to combat climate change issues lies in its ability to help asset owners, or those working on their behalf, to better understand the actual loads and conditions being experienced through the lifecycle of an asset.

The smart use of in-situ sensors and satellite data allows us to identify vulnerabilities early on, making the risks easier to evaluate and alerting asset owners to any potential crisis points long in advance. This enables them to undertake preventative maintenance or strengthening works, or adjust operational schedules, prior to a failure occurring.

As weather patterns become increasingly complex, we can build forecasting tools that model the impact of any scenario on site, from rainfall and windspeed to temperature and humidity. So, for example, what happens if next month’s rainfall is 15% above average? Or if the temperature drops below 15 degrees? With the click of a button, the data can be analysed and acted upon. The days of the TV Weatherperson may be numbered!

As the threat of climate change grows and these risks increase, we need more awareness and insight so that asset owners can better protect their infrastructure, and society – saving money, time and lives.

BKWAI selected for the launch of Deeptech Labs

BKWAI — Mon, 14 Jun 2021 00:00:00 GMT

BKWAI is proud to be one of 5 companies selected to join the first Deeptech Labs (DTL) cohort. DTL is a new post-Seed accelerator from Cambridge and is backed by some of the biggest names in the industry such as Arm, Cambridge Innovation Capital, Martlet Capital, UK tech entrepreneur Ewan Kirk, and the University of Cambridge.

Deeptech Labs’ CEO Miles Kirby notes in an interview with UK Tech News, “We help promising startups by surrounding them with community experts who have been down the same, or a similar path before. Throughout their 13 week journey they are provided with 1-1 mentoring and advice which, I believe, makes a big impact.”

DTL work with an elite company of leaders who have built and scaled deep technology companies. The fact is that new technologies often struggle to be widely deployed, and DTL aims to address this by supporting promising post-Seed startups with funding, mentorship, and access to a huge range of experts and advisors.

We’d also like to wish congratulations to the inspiring companies that complete this first cohort: AutoFill, Circuit Mind, Contilio and Mindtech.

More information about the programme launch can be found on UK Tech News, Business Weekly and CNBC.

BKWAI joins Tech Nation’s next Applied AI cohort

BKWAI — Thu, 15 Oct 2020 00:00:00 GMT

Tech Nation has announced the 32 companies who have been selected to join Applied AI 2.0, their growth programme for tech companies with artificial intelligence at their core – and BKWAI has been chosen to join them!

The programme is designed to champion and support the UK’s leading AI scaleups through peer-to-peer learning and mentoring sessions. Companies who deal with artificial intelligence encounter many of the same niche scaling challenges, and coming together to overcome them and learn from each other is key to the success and advancement of AI tech in the UK.

We are excited to be joining Applied AI 2.0, Tech Nation’s highly specialised growth programme for Artificial Intelligence companies.

Using advanced data science techniques, AI has the potential to solve complex problems fast – and in doing so, this provides great opportunities in being able to free up time for engineers and raise productivity within the construction and built environment sectors. The built environment is a challenging application for such technologies and, together with academic partners, we’re developing methods and analyses that can find something hidden in the data that is not clear to the human eye. Our technologies are steered by the experts and specialists who use them, working alongside subject-matter experts and empowering engineers in their decision-making process.

We’re looking forward to learning and growing in this space, together with some of the most exciting UK companies also joining the cohort.