Considering the importance to any production process, the complexity of almost any industrial operation and the effectiveness of proactive maintenance is hard make short and simple if you can exclude full production stoppage from your calculations. It is obvious that significant impact comes from avoidance of downtime, yet this is only part of the total value.

Understanding the relevance and impact to perform right operations at optimal time on a larger fleet of different industrial machinery requires a thorough look into both direct and indirect costs, the IT-infrastructure, the maintenance processes and above all the strategy that guides all the above. It is for this reason that senior management should take a careful look into the preventive maintenance effectiveness and value they deliver “when nothing happens”.

Some principle questions to prepare:

• What is the impact effectiveness of the preventive team – how do we argue the value that was delivered by actions done at the optimal point?
• Are we ahead of events or scrambling when the alert comes?
• How do we make preventive maintenance decisions? How to make sure it’s not too early “just in case” repair?
• Is my required service and maintenance planning horizon hours, days or months?
• Where does my maintenance operation stand on utilising digital tools to automatise repetitive tasks like inspection routines?

After considering the answers to the above questions, the next question is: What costs do the decisions or operating models carry. What indirect costs are hidden in the way we operate?

Are we following up and arguing how good work our preventive teams and technologies are delivering? – Do we have a standard practise to do aftermath after a repair to evaluate:

Sum of actual repair cost VS potential repair cost

• Production time/device stoppage
• Spare parts + repair services
• Maintenance work + overtime

On top of the delta between actual and potential repair cost, there is the efficiency of daily operations. When focusing on rotating machinery, we claim that many preventive maintenance teams consume a lot of resources on manually inspecting healthy machines (wasted resources) and in parallel find themselves on the situation of colliding priorities due to repairs, measurements and/or other scheduled tasks. This is mainly due we only observe the tip of the iceberg as the preventive process was designed.

The Iceberg model is a great starting point to introduce the depth of the topic. Depending on the size of the operation, the costs get multiplied as the complexity and the sheer volume of transactions increases. Let’s use a concrete example. In the life cycle cost of a pump, the initial cost of the pump represents some 10%-15% of the LCC. The rest include spare parts, maintenance costs and energy – if, we look at the direct costs only. But what lies below the surface? That depends on the maintenance strategy. In a simple exercise, anyone can run through their own operations, ice chunk by ice chunk, cost driver by cost driver. The results of this exercise will reveal very different results, depending on whether your strategy is reactive or proactive.

Modern SaaS-based tools have several benefits for them. We can pick out a few that walk hand in hand with the Ice Berg Model. One, they run in the cloud + edge of the cloud, significantly lowering the need and the cost of IT infrastructure. They are available everywhere anytime, globally – fast to roll out. And so is the data and respective information – centralised by default – available to all interested stakeholders.

Sophisticated tools like Condence.io that have a wide range of methods to detect early symptoms of potential failures, e.g. frequency range in vibration analysis and a large toolkit of analysis tools and algorithms. These will move the detection of potentially failing components earlier, from days to months, and allow for forecasting and tracking, thus enabling a proactive strategy – again globally from one User Interface, moving the detection point earlier and automatic. The ability to track specific parameters translates to time – time for planning. This will help preventive teams to focus on machines and conditions that require attention and plan when to take actions.

Planning results in better resource allocation and cost management as well as budgeting and transparency for the future. Transparency to budding issues, the early symptoms of potential problems results in both less downtime, surprises and increases safety. Failing assets also consume more energy as the performance is not optimal. On top of this, it’s also possible that the root cause of failure creates collateral failures leading to reduced asset life or a replacement. It is generally noted that even only an 8% to 10% increase in temperature over normal working conditions reduces the useful life of an electric motor by half.

Depending on the asset and the nature of the operation, various components of the model apply or carry more weight. An unmanned asset, say a pumping station, is very different from a fully staffed manufacturing facility. They both carry the same issues, just weighted differently. Modern SaaS-based tools work, thanks to their extreme flexibility, in both.

Are you interested to hear more? – Follow up on Linkedin or get in touch with the @distence team. I’d also like to hear your comment and opinion.

References & further readings:

The Computerized Maintenance Management System, An Essential Tool for World Class Maintenance – Michael Wienker, Ken Henderson, Jacques Volkerts

by Janne-Pekka Karttunen
CEO, Distence

While defining the right problem for the traditional study it is vital for Machine learning to have quality data to build on. I have recently heard several times the phrase “your AI is as good as your data”. In predictive maintenance, data often originates from process controllers, designed to optimize the industrial process; therefore, it might not be optimal for predictive maintenance? – McKinsey published an excellent article on this topic, titled “Predictive maintenance: the wrong solution to the right problem in chemicals“.

The authors list four main reasons to justify the title and explain why poor results are delivered: 1) Too little data or too few data points for the system to learn from. 2) Too short time to react; when a prediction is given two days beforehand the preparation of maintenance, which might require a month or more, it provides simply too little time to react. 3) Too little financial impact as processes are planned to be redundant, and 4) Too small savings as the majority of downtime cost comes from planned maintenance. The study considers the chemical industry. While we at Distence can align with many of the arguments, we also see them applicable in many other manufacturing industries.

The first argument raises the point that many professionals are starting to note in public. Mr Tokohimo from Lingköping university noted that “One big hurdle is that the data needs to be collected and stored in a way that allows AI to be applied.“, while David Bell from Ge Digital shared a similar observation in late 2019, “A big challenge in the industrial analytics world is that most machine learning and AI techniques are data-driven. While the industrial world has a lot of data, the type of data needed for analysis from a reliability and maintenance perspective is hard to access because these systems were not designed for this type of analysis“. This is why we at Distence created an open and data-driven platform Condence for the rotating machinery market.

The second argument might originate from the various reasons for failure and from the fact that many of the failures seen in the process data are already “functional failures” which by default lead to short lead times. If, however, the monitoring system is designed for the purpose, in many cases, mechanical failures can be seen months or even a year in advance. For the system to reach this level of accuracy, it is vital that the driver here is again that quality of data and understanding the physics of the problem.

Argument number three is easy to agree within the sense that when the approach is “run to failure” or the financial impact is only considered given the second argument on short lead time. However, I would like to note that if two days would be two months and the data is used to avoid the problem or perform the necessary repairs without a planned maintenance shutdown, avoiding the extra downtime, the financial impact would be substantially higher.

The fourth argument is also very valid in the context of the study. As the McKinsey researchers note later, it is not only converting the unplanned to planned maintenance but to allocate the CAPEX investments as well as operational activities in the most efficient way.

All of the arguments support the fact that the digital journey begins with quality data and applying that right. For rotating machinery, we believe that there is not enough data to build reliable statistical systems delivering long lead times. However, one study made by one of our customers optimized systems on top of the quality data system detected problems in 80% cases before professional’s had caught it in the real situation.

There is a growing need for high accuracy data, stored and labelled in the way that it can be post-processed by machine learning and AI. This type of data necessitates purpose-built metering and systems. This is how we can deliver true value for predictive maintenance with machine learning.

Many of us have moved to remote working from home office to support the fight against the COVID-19. This is usually possible in white-collar positions as the way of working can be adopted relatively easy to the prevailing situation. Remote collaboration tools are the new norm for many.

For those who work in areas where physical presence is required, the challenge is somewhat different. The COVID-19 has highlighted a new challenge – restricted movement or access of people. Even with the high growth rate of automation in industrial manufacturing, many tasks are still perceived as requiring on-site presence. One being condition monitoring of rotating machinery.

We at Distence developed our first concept of remote monitoring and analysis for wind turbines more than a decade ago. The operational requirements we face today were not so different then – “need to perform precision analysis on relevant data, e.g. vibrations, and to access close to real time raw measurement data anytime, anywhere, around the globe”.

“free knowhow intensive work from time and place…”

The objective was to scale the efficiency of highly skilled and scarce resources, e.g. a CAT IV vibration analyst, and free the knowhow intensive work from time and place. This journey has continued, and today the remote services model is more or less a standard in wind power industry. The top drivers for this development were the demand for reliability and cost efficiency of monitoring and repairs. The adoption of remote monitoring was accelerated as cost-efficient technology became available in the early days of the wind power industry.

With the adoption of a new on-line and virtual model, sharing the learnings from global scale remote condition monitoring can eliminate redundancy of maintenance teams on-site. In the condition monitoring of rotating machines, the old process is often dependent on professionals with knowhow and measurement tools travelling from site to site.

Plenty of tacit knowledge locked inside these teams and on the memory chips of their measurement hardware and laptops. Today, however, data acquisition and analysis can be automatized with solutions built on agile and cost-efficient architectures so that the skilled professionals can focus more on knowhow intensive tasks and less on sitting in service vans. Not to mention, all this with increased safety. Remote monitoring reduces radically the impact of sudden unforeseen factors, like sick leaves, pandemics or sudden shortage of manpower.

“Remote monitoring reduces radically
the impact of sudden unforeseen factors…”

Continuous analysis identifies budding issues, failing components and buys the crew time to act with a plan. Information and analysis can be shared without delay with, for example, an external service provider. With remote and automatized condition monitoring, responsibilities can be shared and evened out.

Up until now, monitoring systems have been built on old existing systems, collecting limited data from e.g. PLC’s or they have been built as dedicated point-to-point systems. Often the systems have been built over time with components from several suppliers, resulting in a large number of interfaces, increasing cost and complexity. New solutions are integrated from sensor to interface on modern SaaS-platforms with secure communication from smart terminals to the cloud. Low cost, extremely accurate and agile, these systems are up and running in hours. Globally.

In this modern model, teams working on preventive maintenance can also operate as a virtual team. Distence has partners that are delivering services for several remote monitoring hubs that are independent yet can share resources as well as responsibilities.

What can traditional industries learn from this? For one, remote monitoring equals global monitoring if made free from time and place. Second, the monitoring process and the efficiency and productivity of knowledge intensive jobs can be taken to radically new levels. And three, in challenging times, like the ones we are now living, unnecessary physical on-site presence or travelling and new operational models are valuable.

What do you think? – should this remote working be a norm in condition management profession also. Please reach out to us to discuss.

I have recently noticed that there seems to be many easy-to-deploy condition monitoring products entering the market of vibration monitoring. Many of these are utilising microelectromechanical system (MEMS) accelerometers which are easy and cheap to integrate on the circuit board to incorporate basic vibration measurement features. This accelerometer type, though, is lacking in capability when utilised in condition monitoring vibration analysis or at least in the predictive context. MEMS accelerometers are restricted to well below 10kHz, even as low as 1000Hz, frequency bandwidth which means they are not able to catch essential failure modes like gear mesh frequencies nor bearing fault frequencies, early enough at least. Would you rather have days or months to plan an overhaul for your equipment?

Different Applications Require Different Type of Accelerometers
There are three leading vibration sensing technologies: capacitive MEMS (Micro-Electro-Mechanical Systems), piezoresistive and piezoelectric. They all have their respective use cases, and their abilities in sensing differ accordingly.

From the accelerometer types mentioned above, Piezoelectric accelerometers are the most widely used accelerometer for test and measurement applications due to their wide frequency response and excellent sensitivity. MEMS and piezoresistive type of accelerometers are best suited for either steady-state low frequency (motion) or impulse / high impact (Shock) measurements.

The table below summarises which accelerometer type work in some example applications.

One way to elaborate on the differences in sensing technologies, for example, between IEPE (Integrated Electronics Piezo-Electric) and MEMS, is to compare it to driving your car in the dark with only your low beams on (low-frequency MEMS). If you are driving into trouble, you will still see it eventually, even with low-beams, but it is probably too late to fix the situation on your terms. On the contrary, with proven predictive technology, IEPE (up to 24KHz and beyond) type of sensor, you can sense the smallest cut on a bearing surface, or whatever is in front of you, months in advance.

If you are serious with your condition monitoring via vibration analysis and want to catch or predict failures, you need to be able to source high-frequency vibration. High frequency enables you to capture the necessary failure modes and get a glimpse into the future life of your machinery as it is running today. Last, but not least, you get the required time to react at the time of your choosing.

You can´t use MEMS in serious condition monitoring nor as a tool for predictive maintenance, but if something is better than nothing, you can clamp it on with an “easy to install magnet” without cables and worry about the battery life later.
I wonder though, if it is the installation that is the challenge to be solved in an industrial environment?

Vibration Analysis
Vibration analysis is the most efficient and commonly used condition monitoring method of rotating machinery. It’s the best way to determine the mechanical condition of the device and the possible corrective actions needed to keep it running.
Another good quality of vibration analysis is its versatility. Most machines regularly produce at least some level of vibration, making this type of condition monitoring an excellent choice for companies with many different kinds of devices to monitor.
Vibration monitoring is also an extremely effective method in a predictive sense. Vibration analysis indicates failures in early stages when there is still something to do to prevent the failure from evolving into catastrophic failure. Vibration analysis is a powerful method in predictive condition monitoring; one should just make sure to select the correct tools.

Condition Monitoring with Condence
There are several ways of how condition monitoring can be executed and how modern technologies are utilised. Vibration is one parameter, with Condence, you can combine several different types of metrics, e.g. oil analysis to get a complete view of the health of your machinery. The benefits of condition monitoring and predictive maintenance are endless. We´ve solved a lot of the common challenges in remote condition monitoring, predictive maintenance and IIOT.

The most significant difference between vibration sensing technologies for predictive maintenance is time. When considering professional remote services you need to be able to analyse with enough time to act to the results with well-planned actions instead of the mad scramble of emergency repairs.

Accuracy in measurement gives you time to plan and execute on your terms.

BLOG – Timo Tammisto

Continuing from where we finished in the first part, Decreasing Maintenance Costs, of this blog series on “Why should CFO’s in industrial companies be all over edge computing?”, in this part 2, I want to look at the benefits more from a holistic view by including both some balance sheet line items as well some indirect benefits.

Decreasing maintenance costs is a business case already big enough to get every CFO interested in edge computing and digitalization. This includes not only the industrial assets like rotating machines themselves, but also, and maybe above all, the processes that surround them. These processes, such as Maintenance and Operations or Asset Life Cycle Management, are seriously outdated and have fallen way behind similar processes in other industries where new technology has been used to disrupt old ways of doing things, for example vehicle fleet management or information technology, not to mention the consumer side where our entire lives rotate around various apps and platforms. According to Deloitte, poor maintenance strategies can reduce a plant’s overall productive capacity between 5 and 20%. Recent studies also show that unplanned downtime costs industrial manufacturers an estimated 50 B$ each year.

To see the impact of these new technologies, we can approach the benefits by looking simply at a Du Pont formula or the balance sheet and map the changes from moving to a remote condition monitoring system to the components of each formula. Naturally the impacted line items vary case by case, but at a high level some of these could include the following:

New sales and new business model opportunities from understanding how an asset is really being used, what condition it is in, when to offer service, training or other additional services and when consumables and/or parts that wear need attention – this translates into customer intimacy – having the tools to be truly close to the customer, in a true partnership. There are several examples of new revenue streams, ranging from new services, like training, to selling for example filters, not based on time but by how clogged they are.

Taking it one step further, by having this level of information, it is possible to introduce new business models, sell compressed air as a service, torque as a service and so on. These new business models are built on information, full transparency on the asset generating the service and both its cost- and revenue drivers. Not selling assets, but services, has implications on the balance sheet, cash flow and working capital to single out a few.

If the asset is maintained properly and not run to fault, how does this translate into financial numbers? In addition to the cost implications, outlined in the first part of this blog series, there are several direct areas of impact on the balance sheet. Longer asset life means less capital investments and higher ROA. It also means better resale or residual value. Longer asset life also translates into less interruption to the process as there are less installations of new assets. Higher utilization of assets and respectively higher uptime of the process are naturally the ultimate goal.

These are not just groundless promises. It is simple mathematics. The quality of the information, cost and results of implementing these systems can be easily quantified and verified. And have been, as mentioned in my previous blog on decreasing maintenance costs. The concept is straightforward and easy to understand. We can use our own healthcare as an analogy. Is it better for the body and mind as well as for the wallet to maintain a healthy body or to simply deal with it when sick, missing work and paying for expensive hospital time? With machine health, exactly the same principle of preventative maintenance applies. Only now, with edge-computing coupled with the cloud, the transparency has been taken to a totally new level at a radically lower cost.

From a financial perspective, choosing a condition monitoring system can be tricky – the solutions vary greatly in features and functions as well architecture. The more manual “components” the system and the process supporting it has, the more expensive it is. Using legacy components with less than optimal interfaces and firmware add cost and reduce scalability. The more interfaces there are, not only the more expensive it is, but the more expensive it will be to maintain, grow and add new functionality to. The more integrated it is from user interface (UI) to sensor, the better. Point-to-point systems, systems built around PLC’s and on-site fixed systems can deliver great local value, but they are not scalable – no economies of scale nor scope. The most agile systems can be commissioned in days and scale up without limits.

The new business models and architectures allow a fast and flexible way to implement new systems. When the data transfer relies on mobile technologies, there is no need to penetrate fire walls and integrate the system at the local level to the IT-infrastructure. The sensors can be often attached without stopping the process and due to the fully integrated architecture, commissioning is done in days. Standardized but agile solutions are scalable and can be used to monitor numerous types of assets simply by choosing the respective sensors accordingly.

The business model is simple. The smart units, edge-computing devices, have a unit price and the data transfer, user interface and analysis components are charged on a monthly subscription fee base. Naturally the entire system can be subscribed on a monthly fee as well, including hardware and the cloud offering. Easy to budget and easy to scale up.

Data collection with the intention to only process it further in the cloud will come back to haunt you. Sophisticated vibration monitoring tools, that are the basis for understanding machine health, generate giga bytes of data every day. Moving that amount of data around from numerous assets is prohibitively expensive but also unnecessary. Edge computing is a must here. Embedded algorithms that can be updated and modified from the cloud process the data at the edge, at the asset level, into smart data before transferring only a carefully chosen amount of smart data to the cloud. A scalable, fast-to-deploy, pay-as-you-grow architecture is key.

At the end of the day, it is the information you are interested in, not the infrastructure.

Below are some benefits from Part 1 as a refresher to think about, and to map back to the Balance Sheet and ROI calculations.

As a final thought, digitalization of assets with condition monitoring solutions is easy to pilot. A proof of concept, or POC, can easily demonstrate the financial impact of these new technologies. “Fire bullets, then cannonballs “, as Jim Collins put it. A worthwhile opportunity for all CFO’s to revamp the financials around industrial processes.

1. Transparency. Establishing an 24/7 “umbilical cord” to assets. A source of information to facilitate design and innovation both in technology and business modeling.
2. Continuous Improvement. Through direct access to the assets, how they perform and how they are being used, continuous adjustments are possible.
3. Asset Management. Managing the life cycle of each asset. Knowing what you own and what shape it is in. Extending the life.
4. Optimizing the performance and uptime of assets.
5. Maintenance. Enabling the evolution from reactive to preventive to predictive maintenance. Reducing direct labor, driving times of mobile crews, simplifying logistics, knowing what spare parts are needed etc.
6. Safety and Training. By learning from how the machines are used, serviced and handled, the operator can benefit from focusing training on the relevant areas in order to increase productivity, safety and the life span of the machines.
7. Performance. Thanks to continuous and comparable data – relevant key performance indicators can be built for each asset and process.
8. Modernization. Turning “dumb” machines into intelligent by retrofitting them with intelligent devices.
9. Customer centric design. By understanding how the asset is really used in the customers environment, the R&D organization can better design products from the customer perspective.
10. Data availability. Today the data is trapped in the head of the service engineers or their laptops. Modern tools and the architectures they are built with, make data available for everyone in the organization to benefit from, from O&M to sales to finance.
11. Sustainability and Circularity. Understand the carbon footprint, emissions, energy use and life of an asset.

BLOG – Timo Tammisto

Part 1. – Decreasing Maintenance Costs

New technologies and tools such as digitalization open up radical and concrete new ways for industrial asset owners to totally revamp the financials around industrial assets. These new technologies give asset owners tools to affect both the balance sheet and the P&L with substantial benefits, be they OEM’s or Service and Maintenance Organizations.

In Part 1. of this blog series we will look at decreasing maintenance costs and in Part 2 take a look at a more holistic view.

Digitalization touches on financial elements on both sides of the Balance Sheet as well as both income- and cost sides on the P&L. The impact on the bottom line can be significant. Or, what would you say about payback times measured in the best case in days? Though we have witnessed these cases too at our portfolio company Distence, not all customer cases are that good, but I venture to say, under 12 months in most cases anyway.

We can look at the framework through a very basic but common example. Rotating machines, such as pumps, gears and motors are the basic components of many industrial processes. This is by no means a small asset category. There are an estimated 4 billion rotating machines in service globally with a market value of some €68 trillion. These processes be they energy production, oil and gas, food processing or any other, depend on these components to function. Should they fail, the process stops – and we all know, that is what really costs. In most cases there are tens if not hundreds of them in one single plant, so lots of potential failure points. In one case in Finland, one power plant that has some 20 units, pumps and motors, under constant online monitoring, prevented thanks to the monitoring and warning system the shutdown of the plant twice in one year, resulting in savings over €200.000. Now, if we take just the asset under scrutiny, a pump as an example, some 80-90% of the life cycle cost accumulate after the asset has been commissioned. Not only do these direct costs play a significant role in the LCC but as the life of these assets is very long, 20-30 years, any inefficiencies in any of the cost components compound. Maintenance, spare parts, downtime, energy etc. account for the bulk of the cost. The biggest component in maintenance is labor. Not only is labor expensive, the efficiency is low as well. How much of a workday goes to actual efficient maintenance work and how much into supporting processes? There is a lot that can be done.

On the cost side, some examples of reducing manual work significantly include

• • moving from on-site spot checks to online continuous metering and analysis
  • catching anomalies very early on with the help of 24/7 tools such as vibration measurement before they turn into   downtime for the process
  • optimizing spare parts inventories
  • eliminating driving time impacting both the carbon footprint as well as labor hours
  • turning scheduled maintenance to need based maintenance

Maintenance today is mostly manual, performed by service technicians on site or by crews driving from site to site – spot checks, manual vibration metering, juggling spare parts, troubleshooting. Most of the time of a technician goes into driving to sites, organizing spare parts – anything but the actual fault analysis.

…with modern tools we get another critical asset on our side

– TIME –

time to plan

According to a study by Distence, assuming that a maintenance engineer uses 30% of his or her time on knowledge-based analysis, leaving the other 70% to collecting samples and general tasks, and of all the monitored machines only 10% are found faulty and require action such as repair, we can deduce that only 4.8 hours (3%) of an engineer’s time is actually effective. This operation has not been disrupted for decades. The 80-90% portion of the life cycle cost can be reduced by 60% with new tools and the engineers effective time increased multifold respectively. Reliability of assets can be increased significantly by moving to online monitoring and predictive maintenance. Not only can we save on direct costs, but with modern tools we get another critical asset on our side – TIME – time to plan. Furthermore, we get knowledge, continuous flow of information, material to learn from, understanding of the customer or user of the asset and subsequently a longer asset life. The technician can focus on value adding tasks and leave the data collection to the system. AI, machine learning tools, algorithms process the data and offer deep analysis on the health of the machine automatically, and in the very near future even suggest solutions. All this at a comparable cost of a few visits from a maintenance crew to a site some 30km away.

Cutting costs by rolling out new ways of condition monitoring is easy and fast with new cloud-based solutions that are integrated closely with edge computing smart terminals and sensors and are purpose built. New solutions can be commissioned in days – that’s how close the CFO’s dream is from coming true.

Here are some benefits to think about and to map back to the P&L

1.     Transparency. Establishing an 24/7 “umbilical cord” to assets. A source of information to facilitate design and innovation both in technology and business modeling.
2.     Continuous Improvement. Through direct access to the assets, how they perform and how they are being used, continuous adjustments are possible.
3.     Asset Management. Managing the life cycle of each asset. Knowing what you own and what shape it is in. Extending the life.
4.     Utilization. Optimizing the performance and uptime of assets.
5.     Maintenance. Enabling the evolution from reactive to preventive to predictive maintenance. Reducing direct labor, driving times of mobile crews, simplifying logistics, knowing what spare parts are needed etc.
6.     Safety and Training. By learning from how the machines are used, serviced and handled, the operator can benefit from focusing training on the relevant areas in order to increase productivity, safety and the life span of the machines.
7.     Performance. Thanks to continuous and comparable data – relevant key performance indicators can be built for each asset and process.
8.     Modernization. Turning “dumb” machines into intelligent by retrofitting them with intelligent devices.
9.     Customer centric design. By understanding how the asset is really used in the customers environment, the R&D organization can better design products from the customer perspective.
10.  Data availability. Today the data is trapped in the head of the service engineers or their laptops. Modern tools and the architectures they are built with, make data available for everyone in the organization to benefit from, from O&M to sales to finance.
11.  Sustainability and Circularity. Understand the carbon footprint, emissions, energy use and life of an asset.

In the next blog, I will discuss a somewhat more holistic view and benefits that go beyond cost cutting.

The author is an investor and Board Member at Distence Oy

The benefits of proactive and predictive maintenance are undeniable. There are several methods on how to apply either one. In this article, we use rotating machinery as an example. Most methods fail in accuracy, efficiency, or they are prohibitively too expensive from a modern perspective. Balancing cost, efficiency, and accuracy is sometimes hard at the time of purchase as there are several variables in play. The goal of maintenance is reliability and uptime, being the payback for all actions. The second benefit from monitoring next to the maintenance goals is understanding the use of the asset. How is the asset being used, contributing even more to the equation in the form of end-user satisfaction, training needs as well as feedback to R&D?

Many of the traditional solutions are based on the idea of an on-premise engineer with a relevant skillset to collect the data and to analyze the information. Having people with relevant skills is not given, though. And developing deep knowhow in e.g. vibration analysis is not as easy as one might think. Knowledge also moves along with the person, keeping it in-house is as demanding as retaining good people.

Another approach is to make something simple and low cost. The low-end approach is not a bad idea, as something is better than nothing. These systems can detect that something is wrong for sure, when something happens. The question many times is, what is the process that follows the analysis and detection of a potential fault? Further inspections, root cause analysis, ordering of spare parts, scheduling, shutdown, or a maintenance break? There is a huge benefit from having a lead time, to anticipate and react based on observation, a symptom. The cost of reactive vs. predictive is embedded in the process.

What if there was an option to reduce the cost of traditional monitoring methods significantly, reduce the need for on-premise experts, give lead time to react, and increase the accuracy in our analysis?

To achieve this, accurate measurement data needs to be collected from, e.g., IEPE/ICP vibration sensors. With continuous, statistically valid highly accurate measurements, the operations- and maintenance organization has a continuous pulse on the asset. In the case of rotating machines, as we might be looking at something with low, high, or variable speed, the vibration data can be supplemented by RPM sensors and other process measurements like temperature or pressure to build a more comprehensive view and to see how these values trend and correlate over time.

With this method, depending on the asset and use, the repair cost can be reduced on average by 30%. This is achieved simply by having the insight from the analysis and a heads-up to what repairs to do. This lead time and insight allow for correct and accurate planning, contributing to the many elements of cost in operations and maintenance processes. Moving from manual inspections to automatized monitoring will pay back the system in less than 12 months. The cost elements include less manual work, less field crew drive time, lead time to plan and react, and less downtime, just to name a few. On the other side of the equation, consistent data from assets to R&D for customer-centric design, longer life of assets, circular economy targets, and new business models, combined, can lead to payback much faster. The continuous, accurate data is also a perfect source for machine learning, AI and other applications.

Notably, with a 3-month inspection interval, the theoretical potential failure between inspections is 25%, when with continuous monitoring the corresponding figure is only 0,01% (as continuous monitoring gives close to real-time information). The most significant source of savings is to avoid downtime. In any case, the payback is less than 12 months with higher accuracy.

With the Condence solution, it is possible to use standard “commercial off-the-shelf” sensors, making the system flexible and fast to deploy. The solution is offered with IP67 graded casing to match installation- and environmental requirements in demanding industrial applications. Each device, smart terminal, can have up to 8 vibration sensors and a variety of other measurement points with no need for long cabling – it is easy and fast to install.

The communication uses a built-in GSM-modem while edge computing capabilities, embedded processing of data, allows for optimizing of data transfer volumes. Fully integrated from sensor to user interface, the system is efficient and easy to run, update and maintain.

The architecture results in savings on sensors, installation and data transfer, leading to a significant reduction in both investment and operational costs. Data and knowledge that before used to remain on the premise and often is tacit in nature, is now shared across the entire organization, resulting in both cost savings and new opportunities.

Janne-Pekka Karttunen, CEO

It might be bold to suggest that maintenance is only now reaching the digital age. But for sure we can say that we have not seen the full potential of digital solutions in asset lifecycle management nor in maintenance.

In information technology (IT) one of the core principles has been to store data in one place and in a consistent way. But in maintenance operations, data storage has not been as relevant. Especially when it comes to manual onsite monitoring and data over time.

The focus has been on isolated measurements. Comparing to reference values and subsequently, the data has either not been stored or at best it has been stored on the laptop of the technician.

Traditionally in maintenance, we have not seen data from many technicians. Let alone from many assets and locations aggregated into usable information.

How do we transition to collecting and storing data from assets such as pumps, motors, and gears? How do we manage the data flow and merge information management and processes such as maintenance?

We are yet to discover the full potential of digital solutions in maintenance. How will this transformation proceed? To begin with, we need to understand that there is no one solution that fits all. Having said that, the solutions do share several common components.

We have been delivering solutions to the industry for over 10 years and we have had a front row seat to observe the transformation. Industries and companies that are younger seem to be more agile to transform.

Companies operating in younger industries like wind power, seem to be more mature with digital solutions. They have entered the global market in the digital era. They have built new services on top of new technologies without legacy technologies as a burden.

Also, younger companies with younger management seem to grasp new possibilities without difficulty. This might be oversimplifying, but yet, there is some truth to this too.

The big hype around industry 4.0 has resulted in a large group of start-ups as well as several bigger companies rushing into space, exploiting the new technologies and ways of working that constitute the Industry 4.0…

One can categorize these to two major groups by complexity.

• Simple solutions with a data flow from a simple measurement with a low cost.
• IoT-platforms with data and information management perspective

We have seen that there are possibilities in both. The continuous flow of data from assets poses a big problem for large organizations. Sensors and other sources generate huge amounts of data. Transfer and processing of this data can become prohibitive.

Many of these larger organizations choose to focus on collecting and processing existing data. Data available to them from for example process automation sources.

Startups, on the other hand, focus mainly on one or some very targeted measurements or condition monitoring techniques. This is most likely due to limited resources. For them, it’s natural to focus and be an expert in one thing.

But new technologies such as edge-computing and integrated end-to-end solution architectures open up new possibilities. Agile architectures where the information is processed at the edge and only smart data are passed on.

There are not so many companies focusing on the combination of the two. Companies that tackle both targeted measurements and integrating data into information. Creating a vertically integrated solution.

The ones tackling both approaches face the challenge to be a specialist in several condition monitoring techniques as well as to build an open platform ready for a system-of-systems type of architecture where several datasets from several sources can be combined into one big data.

There is also the rising concern of information security. Data, that did not exist earlier or was on-site and under the control of the internal IT organizations, is now moving to the cloud. And not only to the cloud, but also back to the system, between partners and over the corporate boundaries.

There are several levels of information security to consider in the new world.

From a technical perspective, there is the

• device layer (hardware and software on site)
• security in communication (connectivity)
• application (software offered as service)
• platform (the environment the application is running on)

We have been digesting the topic primarily from a technical perspective. Often it is not the technology that is prohibitive or that slows the transformation down.

The company culture, job security, the fear of the unknown, all come into play quite often. It is not easy to change processes and rethink operations that have been streamlined for many years to reach the current state. The new era can also challenge entire departments and their way of operating. For example the maintenance department.

The new era of digitalization, that has in so many ways changed the world on the consumer side, might lead into a situation in the industrial setting where the entire life cycle service needs a new look.

Rethink the operations from the perspective of the information users and the opportunities digitalization brings to the table.

Even better, with the onset of new business models such as compressed-air-as-a-service or even torque-as-a-service, start the redesign purely from the customer end.

Uninterrupted service under these business models requires new thinking. Ask the question – with what information can I save cost, increase the uptime, find new sales opportunities and better my operational excellence as a whole.

I think this is the whole point – ask the right questions and allow the need to drive the choice of technology.

If we have reached this conclusion, we are ready to overcome the technicalities and merge the old and the new.

Investing time in studying the megatrends of the world, helps think through your own positioning and direction. In a recent Forbes article, out of the 9 technology trends for 2018 outlined, six out of nine had a connecting point with digitalizing industrial assets and the Distence strategy. One, which I left out, the datafication of our lives, also has an interesting angle. In a recent talk at the Nordic IoT Week in Helsinki, Liviu Babitz (Cyborg Nest) had on one of his slides the quote: “I sense, therefore I’m…. MORE”. This quote applies directly in full to the digitalization of assets and the application of edge computing technology in making industrial assets sense more and be more intelligent. The benefit, the “MORE”, is undeniable. Longer life of the asset, higher uptime, more sales, innovative smart services, lower operating cost, just to name a few. All resulting in a higher ROI, not only for the asset, but also for the underlying process and business. Another, perhaps less quantifiable quality of “sensing”, is customer intimacy. Through digitization we get transparency, we are more aware of the assets condition, how it is being used, who is using it and where. Our customers understand how the assets they have delivered to their customers performs the way their customers use the assets. This builds customer intimacy, the ongoing business relationship, regardless of geographical location and increases customer satisfaction and subsequently loyalty and ultimately sales.

Distence has been a part of the above transformation, the various digitalization trends, for over 10 years. Working closely with our customers a small team has been able to deliver solutions to pretty much all corners of the world together with its customers. A big part of this journey has been R&D, perfecting a solution that is reliable, scalable and meets the needs of the customer in each of the chosen verticals we specialize in, with a special focus on demanding environments. The nature of our customers business is such that engineering, and reliability plays a huge part. Mission critical processes set their bars high and it is our job to exceed them.

Reflecting on this journey, fiscal year 2017 and the beginning of 2018 has been a great turning point. Moving from R&D phase to growth stage. The team has more than doubled in size, the board of directors has gotten reinforcement from industry and digitalization experts such as Juha Pankakoski from Konecranes. Krister Hagström, a seasoned executive, joined Distence and opened sales operations in Göteborg, Sweden. Our strong sales promotion efforts, both in Germany and Sweden, are paying off, with the cases and the piloting pipeline increasing multifold from first half of 2017. Panu Virtanen joined Distence as the Operations Manager, also looking over our ISO 9001 certified quality program. Overall, we have been preparing for growth.

On the technology front, the launch of our new Condence-platform delivers through new features and architecture, a reliable and flexible tool for device cloud management. Together with our new smart terminals we have entered, in addition to specific components such as industrial gears and other rotating machines, also equipment builders, OEM, in verticals such as cranes in logistics, paper industry, chemical analysis and more.

Our eyes are now on international growth. The technology is global, that we can state for sure with installations in over 40 countries. The team has deep and wide expertise in growing a business. The competition is tough naturally, but that is good. It means there is a healthy and big market to be served and Distence is well positioned to be among the winners.

In the flattening world, global scale services are a must. If your business’s operations rely on heavy machinery; in the dawn of new technology era, there is no doubt that digitalizing industrial assets would bring significant benefits. Especially, when you own thousands of devices located in a wide geographical range.

This digitizing process is done through adopting edge computing which is integrated into an industrial IoT platform. A complete industrial IoT platform includes hardware with integrated connectivity, embedded software and a cloud-based service with a user interface. Most of general IoT platforms can fulfill those elementary roles in prevalent uses. By adopting only general IoT platform, you might not get full benefits of edge computing. Edge computing is the promising future for IoT: edge computing: the perfect technology solution. We see the need for a flexible IIoT platform with integrated edge computing. We also believe in the system of systems architecture where different IoT platforms can be connected. This way shifting from one general platform to another – or even connect to several platforms is possible. Let’s have a closer look at how these applications can manage to improve your companies’ global service.

Firstly, your company needs a globally-accessible information rather than a team of engineers who are always on call for on-site maintenance. This information need can be fulfilled by using an industrial IoT platform which allows cross-border deployment and assets monitoring in different time zone without local representatives. Now you can save a significant amount of money in labor costs and increase maintenance efficiency.

Secondly, in the edge computing solution, the autonomy of machines is higher and only pre-analyzed data is sent to cloud. This can solve the problem of transferring a large amount of data between remote sites and cloud. Consequently, company can save computer resource and data cost.

Finally, providing the real-time actuation, control, and locations helps your company to detect potential failures or unintended misuse situation of the expensive industrial assets. Hence, unplanned downtime can be reduced, and condition management service stream lined. New services are provided remotely on a global scale which results directly in lengthening your industrial assets’ life cycle.

The digital technology is an enabler for new business models on a global scale. This can be done by redistributing on-site and on-line knowhow services by providing 24/7 customer service globally. From the benefits that industrial IoT platforms and edge computing solution are presented above, organizations might have to reframe their focus to maximize the return on investment and discover new opportunities.

Continue reading to know how Distence helps our customers in transforming their business models to increase transparency, optimization, and control of installed assets.

Distence is an Industrial Internet technology company making industrial assets intelligent and connecting them to business processes with edge computing technology and an IIoT platform.

Our focus is in demanding environments, mission critical industrial assets. We have installations in more than 40 countries worldwide in a very wide range of industry verticals from the mining sector to marine, smart water-applications in the middle of forests and to off-shore environments, just to name few.

There are two main things in common for all of the above. First, industrial applications have generally a very long lifetime, some going beyond 30 years. The second common to all, is the requirement to manage such lifetime in challenging operational environments: cold, heat, dust, vibrations, humidity, sea water and EMC (Electro Magnetic Compatibility). Today we have installations where our Customers have used our Smart Terminals for almost 10 years in in 24/7 use.

We at Distence have the experience to develop hardware to meet these kind of needs. We continuously develop ourselves and the ways we do our internal R&D-testing to verify the products truly meet the requirements.

Attached video is taken from our lab and provides a brief look into one aspect of system testing we do.

https://www.youtube.com/watch?v=Xva2bYwP_Bw 

The benefits from digitizing industrial assets
Information has no boundaries – it can benefit all functions

As the competitive landscape with the digital revolution is changing and the pace at which it changes accelerates, companies must also be ready to a multitude of potential outcomes – flexibility to react and to adjust will be key. Choosing one path simply is not possible. You must be ready to change course fast to survive. Alarmingly, according to a PWC-study, most companies, up to 70%, were focusing mostly on efficiency and cost cutting. Good examples being predictive maintenance, energy efficiency and remote monitoring. According to this study, the most balanced approach came from the Chinese, who also had the above-mentioned cost and efficiency as targets, but who also had the highest count in new business development as a benefit from digitalization.

Flexibility is one of the cornerstones of the Distributed Intelligence concept. What do we mean by flexibility and how does that differentiate the approach? We see flexibility in two parts, intertwined together. First, flexibility in the way we work. Starting with a question about the customers’ business needs, with what information and how can they benefit from the information coming from digitalizing assets. Only after that need is defined, can we move to the technology. Even though this might sound as a common approach, identifying business needs first, this is not common, or genuinely true, in an industry combining hardware and software for mission critical assets, very much driven by technology, a point-to-point solution, a plug to a need.

To be truly flexible and business driven, the solution has to be flexible enough to accommodate the customers’ needs to a degree, where they can build competitive advantage to their offering. However, as said in the beginning of this article, the rate of change is accelerating, so the business needs of today are not those of tomorrow. At the same time as our customers are launching new services using their new digital innovations, they need to be ready to react to the respective offerings from their competition. Room for maneuvering will be needed, one way of the other. If today the pressing business need is cost cuts and efficiency, tomorrow it will be business development, safety or quality. Too often still today, especially large companies have deep silos with thick walls, well protected budgets and territorial wars. But data knows no boundaries. From data you get information, and information is power.

As an initial starting point, remote monitoring might get the budget, after all it does cut cost and increase uptime. It also has a ROI that is easy to calculate. But if the asset or process is viewed across functions, and the technology harnessed to serve functions across the organization, the ROI could be multifold. The CFO is for sure interested in energy consumption, asset life and depending on the asset, utilization levels and possible revenue streams. HR can be interested in who is using the machine and do they have the necessary training or certification to use the asset. Sales is always interested in selling a new unit, new filters or other consumables or in understanding how the customer is using the unit – how could you increase customer satisfaction. The possibilities are never restricted to the needs of just one function. Technology must be built to support the above opportunities. The technology, both the hardware and the software, must also be dynamic, active. 2-way traffic, meaning not only monitoring and data collection, but also control, optimization and management. The asset must be made smart, if possible autonomous. Software updates are important, for both the operating system as well as the functions and features. The solution, from the smart terminal, edge-computing solution, to the platform must be able evolve as the world around the asset changes. Flexibility is king.

Distence participated in the Resource Waste Management (RWM) trade show in Birmingham UK from 13^th to 15^th of September. Parallel to the main event there were additional events on Energy, Water and Renewables sectors. This was the first time for Distence at the show and we were part of Finpro’s stand together with five other Finnish cleantech and IoT related technology and service providers. It was a great opportunity to get an overview of the waste management sector along with its machinery, equipment and service needs and offerings.

During the show we captured few key themes, one being the increased focus on recycling and waste to energy type of technologies and solutions and the second being waste collection optimization. Naturally we were seeking for business opportunities for Distence’s solutions in these themes. Just by looking at the machinery such as compactors and balers in the waste management value chain, there were versions with very low level of automation to sophisticated and fully automated units.

In many of the solutions, however, no remote monitoring or management functionality was implemented. With Distence’s intelligent edge computing and cloud solution, remote monitoring of fill level, hydraulic pressures and error messages can be implemented, just to mention a few parameters. This could be further expanded to localized automated actions or remote control of key parameters to enable automation, at a decent cost level, in lower end machinery from the warmth and coziness of an office. According to our discussions with several waste machinery makers, just by reducing the visits for on-site service could bring hundreds of pounds of savings per machine on a yearly level. This may not sound a lot, but with hundreds or even thousands of e.g. compactors in the field, it adds up to significant savings. All this in addition to the avoidance of unscheduled downtime in the use of the asset. Further more the possibility for manufacturers to collect information on the use of their product to take back to product development, sales and quality departments, increases the number of information users and the ROI from the investment. For mobile machinery the information, made available thru the solution’s fill level information together with GPS location data, can easily be used for waste collection route optimization for increased efficiency and cost savings.

The edge computing technology and remote management solutions are ready. It is now up to the businesses and operations to capitalize on this available competitive advantage.

More information about Distence and the applications at www.distence.fi or Kalle Nurminen (+358 40 538 2216).

I believe it is not wrong to say that in every company digitalization is discussed, in one form or another. That is, how to capture

the possibilities that new technologies bring with them and how to develop the business. And this is all good.

In my own work I have however realized that too many companies end up gravitating too fast towards technology when exploring the possibilities of digitalization. This way of working leads alarmingly often to the conclusion, that IoT is about technology and that the knowhow in this technology must be in-house. As soon as we have a platform and the specialists, we can get to work.

When choosing this road, an IoT technology department is born with a mission to build and maintain the best in class technology. Investments into IT, software, licenses and equipment are done according to the traditional company purchasing process, all well prepared and planned and using the good old purchasing department bidding process. You can see and feel the energy, there is so much to do and the excitement around the new technology is almost tangible.

Who would not want to be part of inventing the wheel?

The focus and fuss around the technology consumes however enormous amounts of critical resources, both management focus and time. All this on something for which solutions already exist. This start-up phase takes easily a year, which is the core of the problem. Meanwhile the competition is already far down the road implementing the learnings and rolling out the next generation products and services.

An agile organization focuses on thinking, what is it that we are trying to achieve with digitalization. They explore, pilot and iterate. Success does not result from acquiring technology and resources, but from the intelligent application of it. This is why most of the resources should focus on the latter. A partner with the appropriate technology and tools can take care of the rest.

I firmly believe, that regardless of the hype around IoT, it does not differ, from the perspective of technology and operations, at all from traditional IT. An outsourced IoT-service frees the organization to what it knows and does best. Outsourcing the IoT-technology and the operations keeps the focus on the business and its development.

It is up to both the IT-partner and the IoT-partner to ensure that the service and the underlying technologies develop continuously. A good partner offers ideas, support and allows for the customer to leverage the partners wide experience and knowhow.

On every strategy course the classic question, ”make or buy”, is presented and discussed. This question is still crucial to the success of any company.

Markku Hollström
Vice president responsible for Elisa’s IoT projects
ELISA Oyj

Guest Blog
Jukka Toukonen
Service Specialist
ABB Oy

The majority of the life cycle costs of an electrical motor comes from energy consumption. The purchase price represents some 2% of TCO (Total Cost of Ownership) of a motor with a life of some 20-40 years. Electrical motors have parts that wear down (e.g. bearings) in addition to which dirt and mechanical and thermal strain wear down the motors. In-shop service is recommended depending on the type of use (type of load, number of starts, environment) every 10-15 years. Even coiling might have to be done 1-2 times during the life of a motor, this also is done in most cases at the shop. The share of life cycle costs of maintenance represents some 10% and coiling up to 50%of the purchase price.

In addition to the maintenance costs also preventative work must be included in the total costs, this includes e.g. lubrication, condition monitoring, cleaning etc.) The costs of preventative measures can be significant over the life of the motor.

As a whole, the maintenance cost can equal the purchase price when calculated over the life of an asset. This sum can easily grow to tens or even a hundred thousand euros over the life for bigger (> 100kW) motors. Condition based maintenance aims to minimize these costs while maximizing the up time.

A scheduled maintenance program is easier to budget, but it falls short on cost when compared to condition based maintenance. A failure caused by aging or wear come seldom as a surprise of the asset is monitored regularly. The challenge with periodic monitoring is, however, in keeping the conditions same and te measurement results comparable. The conditions, load, use all vary). With continuous monitoring it is possible to not only spot ong term trends but also short term fluctuations, with which is it easier to predict maintenance needs and the impact the maintenance has. Very often the requirements for the motor change over the life of the motor and might differ greatly from the purchase moment of the asset.

One of the big benefits of continuous monitoring is to get to the core of the biggest cost component – energy consumption. For example understanding temperature tells a great deal about load and loss. Vibration monitoring can lead to the contributors to loss such as resonance, imbalance, misalignment, lubrication and friction. Vibration consumes always energy, which is away from the actual purpose of the motor – generating torque. Continuous monitoring also tells how the motor is being used. From the results, one can read how well the motor serves the production process, and it can also give indication to the optimization of the entire process.

Implementing continuous monitoring is technically easy and the cost is reasonable. Ready solutions and components are readily available. Implementing the technology is easy, the question is who can set the correct parameters and interpret the results.

ABB Motors and Generators Service offers condition monitoring services, where the customer can purchase a full service monitoring and analytics service. ABB will deliver and install the system,, configure the parameters and if desired take care of the monitoring and interpretation of the results.

The entire solution can be delivered as a service, with one monthly fee. The system can also be installed without interrupting the process.

This type of a monitoring as a service does not require personnell or equipment investments, nor does it require additional training.

I believe that in the future new services will emerge as an alternative to buying a motor. There will be service that take into account better the entire life cycle costs, for example, buying a guaranteed torque with a fixed monthly fee. In a business model like this it is imperative that the service provider has 24/7 information from the motor and its environment and use. As an intermediary step to this kind of a torque-as-a-service business model could be a torque-guarantee from the equipment provider in exchange for a 24/7 access to the same information.

Guest Blog
Pekka Savolainen
Master of Things
Affecto Finland Oy

During the last year I have had the privilege to discuss with several large and small companies about digitalization, internet of things, industrial internet or advanced analytics – with name of the topic varying from company to company. In these discussions, the biggest challenge has often been in defining the meaning of the topic for company’s business.

Unfortunately there is no single answer. There are plenty of success stories from all corners of the world on how others have realized the possibilities of digitalization. But the significance of digitalization must be realized in every organization separately. Improving or optimizing current operations or increasing safety might be essential to some. For others, asking what is the value delivered to the customer and how could we serve the customer better is more important. Perhaps digitalisation changes company’s business model by enabling them to sell their products to the customer as a service? Or perhaps the company can respond to the customer needs better by doing closer co-operation with its current or new partners. Reading case studies from other companies can trigger the imagination, but copying solutions from other companies seldom work as such.

Time spent by creative people often results in new insights and directions stating how things could work in the future in totally new ways. These visions are important to establish a goal to aim at. However, too often the innovative ideas make nice Powerpoint presentations, but nothing more. Following the Lean-thinking, even a small improvement in production brings more value to the customer than a grand idea written on a Powerpoint.

I recommend to challenge employees, customers or partners by asking them to identify a specific thing, which when solved or improved, would deliver an immediate benefit to the customer and the business. Usually a small group of people can come up with 5-10 ideas immediately. From these we can pick one and test its feasibility. In the case of Affecto, the experimenting often revolves around data – we can take for example a static set of data from the customer and study whether the data answers our customer’s business question or not. Or we can build a model based on the data to predict a phenomenon that is interesting to our customer and estimate how accurately the model works. When data analysis has verified what we can achieve based on the data at hand, it is easier to decide whether the idea is worth investing and risks are significantly smaller. Alternatively, the result of the analysis may state that the currently collected data is not sufficient for intended use. Failure has value in itself, as it directs you to focus first on something else. This is why it is important that the experiments are carried out small, with low cost and risk.

The good news is that enabling technology is here and supports fast prototyping. For example, the Oliotalo ORM technology is an easy way to create connected assets and start benefitting from analytics. The other good news is that the cost of such experimenting is low: what was not affordable yesterday, will be tomorrow. Let’s roll up the sleeves and start experimenting!