Industrial IoT: Looking backward and forward at BLAST 2017 in Rome

At the recent BLAST conference in Rome, Italy, Alessandro Bassi, president of IoT Italy, moderated a panel about IoT in Industry.  Sergio Gusmeroli, Research Coordinator at the Politecnico School of Management in Milano spoke on the importance of IoT for Industry.  I introduced some Industrial IoT applications, and Maurizio Griva, Senior Manager at Reply (a system integrator) introduced the “self-aware factory.”

Technology alone is worthless.  We need value-oriented transformation of industry

Sergio Gusmeroli, Research Coordinator at the Politecnico School of Management in Milano (www.dig.polimi.it) set the stage by stressing the importance of IoT for Industry:  Smart Manufacturing makes up about one quarter of IoT-based innovation and will have the highest growth of IoT subsections.   Manufacturing is also the industry sector that spends the most on data-driven innovation.  Mr. Gusmeroli introduced the US and European manufacturing initiatives, including Industrie 4.0 and its variants. The European Commission helped drive these and Commissioner Oettinger’s digitization initiative reinforced the initiative.

Mr. Gusmeroli referred to the US and EU policies and initiatives to create growth through manufacturing and cited a few strong statements that illustrate that digitization and software in themselves are not sufficient to transform the industry in a value-generating manner.   What is needed, Mr. Gusmeroli argued, is a maturity model that helps transforming industry in six main domains with appropriate, each with a few key elements and corresponding KPIs:

  1. Products:  IoT-enabled products and services
  2. Production systems: Green and sustainable, near-shored, attractive workplaces
  3. Performance: real-time IoT architectures, edge and fog computing, and real-time analytics
  4. People: Human-centered production, employing younger workers with innate digital skills
  5. Platforms: open source, standards-based, interoperable next-generation enterprise systems
  6. Partners: non-hierarchical collaboration; open, “living” labs and startups; and digital innovation hubs.

ARC also observes the emerging use of non-traditional, non-financial KPIs.  We will be following Politecnico’s publications in this domain with great interest.

IIoT will evolve from add-on applications to fully cloud based.  Humans need to transform too

Representing ARC Advisory Group (www.arcweb.com), I introduced industrial applications of IoT, from mature to visionary.  IT- and OT-based technologies have not been applied to their full potential, and therefore “Industrie 3.0” is not over, while “Industrie 4.0” is being implemented.  “Add-on IIoT” applications are most common today.  These cloud-based applications exploit connected sensors in production environments to provide additional insight through analytics (asset-analytics, quality analytics, process, and/or supply-chain analytics). More complex analytics applications involve several analytics domains.

In parallel, the creation of collaborative work environments in which video, voice, augmented reality, and analytics support distributed teams. In some companies, human transformation goes hand-in-hand with digital transformation.  These companies create (more) autonomous teams, favor creativity and collaboration, and provide environments in which people can flourish.  The economic results of those purely human-centered initiatives can improve profit and turnover.  Finally, the trend started with virtualization of IT solutions in manufacturing, will probably lead to an increasing portion of IT and OT in the cloud.  I expect the complexity of IT, OT, and IoT to increase as well as the scope of associated optimization.   When these trends continue to develop, they may ultimately lead to a fully cloud-based IT, OT, and IoT environment. This would make it feasible to execute any product design or recipe from anywhere in the world on any available production equipment.

Requirements for sensors for IIoT and how to build the self- aware factory

Maurizio Griva, Senior Manager at Reply (www.reply.com), a system integrator in the discrete manufacturing domain, introduced the “self-aware factory.”  According to Mr. Griva, this includes data ingestion from the edge; data aggregation at the edge and in the cloud; and predictive maintenance, anomaly detection, and automated incident tracking.   IIoT are “the ears and eyes” of the plant floor, he said.  The challenges that Reply observes are:

  1. The availability of production-grade sensors, with precision and reliability adapted to industrial application
  2. State-sensing sensors, that can pre-process signals for downstream model-based interpretation of the values it produces.
  3. Context-aware sensors: when pure process measurements are enriched with context measurements such as light, temperature or humidity, people and machines can make better quality decisions
  4. Multi-sensing sensors, able to acquire several physical quantities, such as shock, strain and vibration
  5. Consumer electronic sensors.  In parallel with industry grade sensors, consumer electronics sensors, for example smart phone cameras can provide a cost-effective context information

Mr. Griva mentioned the following applications and use cases:

  1. equipment health monitoring
  2. product and equipment tracking and localization
  3. automatic incident creation from field data
  4. energy consumption/production management
  5. warehouse optimization through AGVs
  6. predictive maintenance

Griva recommended to “start with what you have” and build on that in four steps: monitor, apply monitoring information for maintenance, profile the full plant, and turn these observations into predictions.

Conclusion

In these talks, this analyst observed a very consistent vision of the current state of and potential for IIoT.  The common challenge identified is gaining the appropriate awareness in small-and medium-size companies, since acceleration of IIoT adoption will only be possible with increased awareness.  This acceleration would likely have a significant impact on economic growth, jobs, and sustainability.  A good maturity model for industry (such as the one Politecnico is working on) will be critical; technology availability is probably the least scare resource in the mix.

Comments

  1. I personally agree that apart from the digital technology, the work procedures shall be rewritten around the new technology; a “check the software first” mindset. That is, you “google” the pump before you go to inspect. Indeed, being ABLE TO change the work practices is the ultimate objective of digital transformation. See this essay how plants do it.
    https://www.linkedin.com/pulse/digital-transformation-best-practice-maintenance-jonas-berge

    Automation system based on 4-20 mA and other analog and on-off signals are not very exciting to work in. By going digital many new enchanted products are available and even more possible for the future: https://www.linkedin.com/pulse/saving-time-magic-its-method-jonas-berge

    I also agree plants need real-time digital networks. Not just for IoT, but for local edge/fog computing and process control. http://www.fieldbus.org/images/stories/technology/aboutthetechology/overview/fieldbus_brochure.pdf

    Younger workers will want to be digital workers. A tablet instead of a test pen or a multimeter. This requires devices which are digital, networked, and can be “Googled”.

    Personally I also believe “platform” middleware will go towards standard-based open solution. That is, not a platform technology owned by a single company but an open technology is owned and managed by a manufacturer-independent open multi-vendor organization like OPC Foundation etc. where members have joint input on the direction of the technology.

    Industrial grade sensors which are reliable and have high accuracy are available

    These sensors are installed on equipment to pick up vibration, acoustics, and corrosion in addition to the usual pressure, temperature, level, and flow etc. to feed into equipment analytics software

    One of the many reasons why digital sensors (sensors with digital networking instead of 4-20 mA signal) are used is not only to reduce wiring and simplify installation, but to carry multiple variable from multiple sensors in the same transmitter.