MICROSOFT: The Hannover Messe Report

We mentioned in Part I earlier how Hannover Messe is emerging as an Industrial-IoT conference at ts heart, with an equipment showcase included. The size, organization and content quality of HM17 has grown to dwarf most IoT conferences around the world. As always, ArcInsight Partners appreciates the time companies spend with us on 1:1 conversations and analyst briefings during conferences.

We have mentioned earlier how Hannover Messe is emerging as an Industrial-IoT conference at ts heart, with an equipment showcase included. The size, organization and content quality of HM17 has grown to dwarf most IoT conferences around the world.

For Part-II of the Hannover Messe series we chose to focus on MICROSOFT – a company that has perfected the art of showcasing industrial-IoT credentials at major conferences. The company has no real origins in IoT nor any native pedigree there (its an enterprise software provider). Yet it enjoys one of the best “mindshares” among its peers vying for a place on the IoT bandwagon. Microsoft has made its way to the green box in our Industrial-IoT solutions map.

Instead of placing itself as the cynosure of its own paid-for exhibit areas, the company cleverly herds a group of companies within its client and partner ecosystems (all carefully selected around a specific theme), then unleashes them to tell a value delivery story about themselves or about their own customers, while mentioning (subtly ofcourse) how they leveraged Microsoft’s core IoT offerings – the Azure cloud-services, machine-learning algorithms, the Holo Lens mixed reality tool, among others.

To an astute observer, Microsoft’s marketing has successfully made the subtle (but significant) shift from a product advertisement strategy (“it’s all about me”) to a more product-placement strategy (“its all about you…with me in it”). Smart !.

ArcInsight Partners believe this is the kind of outcomes-driven story every serious player in the industrial-IoT space must go to market with. IIoT is a team sport and is also a long gestation game. We have written about this on previous articles such as “Monetizing The Value Of Industrial-IoT“and “IoT Has Moved To Its Next Phase Of Evolution“. While we see the business case for some of Azure IoT services & tools as quite clear for use as industrial-IoT infrastructure elements, others are at best unclear. More on that in ArcInsight Partners’ own observations below.

One Microsoft ecosystem partner that appears to have created monetizable value from a manufacturing context is the Swedish company Tetra Pak. While not quite a household name in America, or for that manner elsewhere in the world (it certainly deserves to be) given the number of interactions consumers have with its products everyday, and in nearly part of the world – The humble Milk-Carton being one. Last year Tetra Pak sold 188 billion packages and produced $12.2 billion in net sales.

Few can possibly realize (and ever fewer bother to think about) how design parameters of a seemingly ordinary looking paper carton allows milk to stay fresh, consumable and free of harmful micro-organisms through its entire shelf-life. Tetra Pak plays a key role in making sterile and aseptic packaging the foods industry depends critically on for shipping its products. In the aseptic process, short heating and cooling times (275 degrees Fahrenheit for up to 15 seconds) ensure food is absent of harmful microorganisms while retaining its taste and nutritional benefits. The food is then packed in pre-sterilized cartons. An unopened aseptic package can be stored without refrigeration or preservatives for up to six months.

Their key success factor – “The Better We Can Ensure Packaging Integrity, The Better We Can Ensure Food Safety. This Results In A More Perfectly Shaped Pack.” When creating 190 billion packs consistently, Tetra Pak’s systems help the quality and look and fit of the pack.

Ensuring this requires maintaining the integrity of its manufacturing process, critical components humming along exactly as they were designed for. While that assumes preempting any possibility of equipment breakdown through good conditioning, it also requires that system flow is not interrupted too often for inspections to ensure smooth operations. Interruptions are also costly from a production loss standpoint, not to mention unnecessary in many instances. (We wrote about this in Monetizing The Value Of Industrial-IoT )

When one part fails on a milk-packaging line, it can take the entire operation offline for days, leaving a plant unable to store thousands of soon-to-spoil gallons of unprocessed milk. And there’s no stopping cows on the farm or the milk that must be harvested. Do it too soon, a plant may unnecessarily swap out expensive parts well before they’re worn, adding costs. Do it too late, a plant may have to wait days to repair a machine and restart operations. “You could then have a very long down time because the spare parts may not exist in a country where the customer is. It could take days to get parts there. It can be the difference between a four-hour, planned stop – where they might stop anyway to clean – and a four-to-five day loss of production where they can’t use any equipment at all.”

Minimizing vibrations in these machines allow to company to keep package quality within acceptable tolerance levels. The better the packaging shape and condition remains, the better the packs are going to look and perform. Naturally performances drift over time and anticipating such drifts make the difference between detecting failures. The company has begun to offer predictive maintenance services to its food and beverage manufacturer clients in more than 175 countries, through reducing the amount of downtime and unexpected outages. The company has placed sensors on some carton-filling equipment, allowing global experts to analyze, in real time, data patterns in those factories against data from more than 5,000 other packaging lines.

Data that’s gleaned from connected machines gives company technicians an operational snapshot across more than 5,000 Tetra Pak packaging machines in use. That information helps service engineers swiftly answer critical questions, such as: When did a specific failure happen and how long did it last? What was that machine’s temperature at the time? What was the pressure?

Tetra Pak started rolling out HoloLens mixed reality headsets to Tetra Pak on-site service engineers – who are now able to work with the company’s global experts that specialize in production line problems. A great way to enable collaboration and seamless joining of business processes to deliver value.

As Johan Nilsson (Vice President Tetra Pak Services) puts so eloquently:   “There Is Consolidation In The Foods Manufacturing Industry, Like Many Industries. The Bigger The Plants Become, The Higher The Cost Of Unpredictable Failure Becomes Per Minute. And The Higher The Cost Of Unpredictable Failure Becomes, The Higher Is The Value Of These (Predictive Maintenance Analytics & Mixed Reality) Tools To Support Field Maintenance.



We see a great business case for Azure IoT cloud and machine-learning services and is quite clear for an infrastructural element in industrial-IoT. However mixed reality tools (such as Holo Lens, an excellent design innovation one might add) have been used effectively only in a handful of use-cases categories so far. The rest have come across as merely gimmicky industrial mockups, of value only as sales-demo tools. Clearly Tetra Pak has presented a very appropriate use-case in effectively using Holo Lens for predictive maintenance and field service management – two of the most popular ones.

Not Every Industrial-IoT Use-Case Lends Itself Well To Use Of Virtual/Augmented/Mixed Reality. There Is A Place In IIoT Applications Design For Common-Sense & Design-Thinking Methodologies For Investment Decisions. Doing So Could Make The Difference Between Returning A Big Leap In Operator-Effectiveness, Versus An Unnecessary Investment In Gimmicks.

Conditions inside the fermentation-tank of a beer manufacturer or wine-maker are far less dynamic to require high resolution sensor data sampling, and can easily be monitored using plant-floor level HMI or remotely from control room display-panels of a DCS system. The worst downside of critical process-parameter deviation may be the failure of an entire batch of (work-in-process) premium beer intermediates, which may lead either to a downright disposal (full batch cost write-off) or perhaps its re-purposing for a lesser-value beer product (only a partial inventory loss). In either case, remediation simply involves draining the tank, cleaning and a reboot of the process, not a significant damage. The impact to the entire shift’s beer production while costly, is relatively low for the manufacturer.

In contrast, a fast moving gas turbine running at peak load has hundreds of moving parts each of which could become the cause of catastrophic failure without high-resolution performance monitoring preferably with trending and predictive capabilities. An unmonitored deviation of operating-parameters requires better early-warning system. A catastrophic failure in this situation not only “offlines” the equipment for days, its remediation involves time to diagnose root causes, scheduling service technician visits to site, ordering and shipping parts, performance testing and startup proccesses, before finally bringing it back online. Not counting the impact to production losses. The total cost therefore goes way beyond a production loss on a shift.

We have no doubt revenue will continue to grow for Microsoft’s IoT cloud and machine learning services. While cloud services business is a frontrunner in IIoT industry mindshare, the company still lags industry leader Amazon’s AWS by a wide margin in revenues (estimated at $13Bn Vs Microsoft’s rumored $2.5Bn. Google is behind. IBM/HP are practically invisible in cloud.).

Augmented-Reality (AR) & Mixed Reality (MR) Remain Unclear Value Propositions For Industrial-IoT, Except In Well Analyzed & Well-Designed Use-Cases.

There are good use-cases in discrete manufacturing and field maintenance with tremendous opportunities to impact business process delivery, leading to clear defensible value-propositions. There seem relatively little value for leveraging Holo Lens inside core continuous-process environments and even lesser for slower moving batch-process environments. For Microsoft, bundling all of its available IIoT targeted offerings does make good business sense. Over the long haul however, it’s monetizable value creation at the business-end of use-case applications that drive their popularity & longevity. As always, we are excited at opportunities to carry out deeper examination of your operating industrial-IoT use-cases (using ArcInsight Partners’ structured proprietary methodology).