Recipes for Success: A Culinary Guide for Advanced Industrial Analytics

In a recent blog, I introduced and defined the technology category of Advanced Industrial Analytics (AIA) and provided an update on the upcoming solution selection matrix. In that post, I also explained the three types of AIA providers and how they are positioned within the Industrial Transformation Reference Architecture.

This blog post expands on that and describes three ways for industrial companies to consume Advanced Industrial Analytics. It lists the qualified vendors assessed in each of the three categories and highlights key characteristics in the context of three popular dining options.

1. Taking an à la carte approach: AIA by Industrial Application Providers

   The first category, which includes Advanced Industrial Analytics (AIA) solutions offered by application providers, can be compared to an à la carte menu that offers a variety of individually priced dishes that can be ordered separately. Customers have the freedom and flexibility to pick and choose individual items based on their personal preferences, constraints, and restrictions or even order separate items from multiple vendors.

   Similarly, pureplay AIA application providers offer software applications that can be purchased or licensed without many constraints from external restrictions, such as compliance with IT policies, cloud infrastructure, etc. Ideally, these applications can be accessed from either a web browser or a local instance that is not tied to a particular platform, infrastructure, or ecosystem.

   Since it does not require significant IT involvement, lengthy implementation, and custom development, these applications can be licensed relatively fast and up and running quickly, similar to ordering off an à la carte menu. However, while this approach has more flexibility, readers should know that each item serves its purpose. Just like with an à la carte menu, it is upon the consumer to create their meal of appetizers, entrée, desserts, and drinks, it is upon companies to pick and choose individual analytics applications for different business needs.

   Additionally, no one particular vendor in this category offers 100% of a typical manufacturer’s analytics needs, so taking this approach for analytics means that companies will need to choose a couple of AIA applications (on average) to provide a complete set of analytics capabilities based on their business objectives, use cases, user personas, and digital maturity.

   The Solution Selection Matrix that covers the Advanced Industrial Analytics application providers will assess the following vendors:

   1. • • Augury

        • Canvass AI

        • GE Digital

        • Oden Technologies

        • Quartic.ai

        • Seeq

        • TrendMiner

2. Choosing the breakfast buffet: AIA by Industrial Data Platforms
   

   The second category, which includes Advanced Industrial Analytics (AIA) capabilities offered by Industrial Data Platforms, is comparable to a buffet-style menu that contains a predefined set of items. Just like a breakfast buffet contains an assortment of food choices, Industrial Data Platforms provide data connectivity tools, multiple applications, no-code/low-code development environments, etc., that collectively make a full-fledged analytics solution.

   The principal value proposition of an Industrial Data Platform is that it doesn’t just provide a solitary application but brings together multiple sources of data to a homogenous environment and creates an enterprise-level data model (digital twin) of assets, processes, or material flow. Sitting on top of this data layer, commercially off-the-shelf applications provide advanced analytics across several use cases.

   Additionally, just like one could supplement a buffet with external items, like a separate omelet bar or a cup of coffee from the outside, the Industrial Data Platform enables easy connectivity to external applications, thereby enabling a Best-of-Breed architecture where external applications can feed off contextualized data from the data platform and its own set of applications.

   The approach here, then, is not picking individual items off a menu but partnering with a vendor to build a holistic data layer, provide analytics for multiple use cases, and enable efficient collaboration with external applications. While this approach provides a more comprehensive set of analytics tools, the downside is that its data connectivity and application enablement require more IT involvement than a SaaS product, leading to a longer time-to-value than the à la carte approach.

   However, like most analogies, this one only holds its ground to an extent and should be taken with a generous grain of salt. For example, unlike a buffet menu where you pay a standard price regardless of what you consume, the Industrial Data Platforms provide multiple consumption and subscription-based pricing models based on the size and scope of the deal.

   The LNS Research Solution Selection Matrix that covers the Industrial Data Platform providers will assess the following vendors:

   • • Braincube

     • Cognite

     • Hitachi Vantara

     • Machine Metrics

     • Rockwell Automation Data Mosaix

     • SightMachine

     • ThinkIQ

     • Twin Thread

3. Exploring a tasting menu: AIA by Industrial Application Platforms
   

   Finally, the third category of AIA by Industrial Application Platforms can be compared to a specialized tasting menu with a preselected list of drinks, appetizers, entrées, and desserts.

   It is important to note that the application platforms have a data platform component nested within them, as shown in the IX Reference Architecture. This means that while the application platforms offer data platform capabilities like connectivity tools, DataOps software, data modeling, and applications, they also go beyond a data platform and provide a more sophisticated no-code/low-code development environment, DevOps to manage application deployment, and a marketplace of COTS or custom-built applications for customers to license and share easily.

   However, their data platform capabilities typically fall short in terms of depth of functionality compared to pure-play data platforms. For example, their connectivity might require the help of third-party partners, their data modeling capabilities might not be as robust as a pure-play data platform, etc. The application platforms’ main value proposition lies in the ability to enable customers to quickly design, build, and deploy (COTS and custom-built) applications at scale.

   Similar to a tasting menu where you don’t go just for the food but also for the ambiance, the service, and the overall experience, the application platform approach should be assessed not just by the applications but also by the connectivity, contextualization, data model, development environments, implementation services, and partner network.

   Considering these additional capabilities of an application platform is important since this approach requires a substantial investment of time, IT involvement, and commitment to one particular vendor. Companies taking this approach are hence advised to not just look for a software/platform vendor but a trusted partner they can commit to for the long term and work with to build this environment of homogenous analytics solutions.

   The LNS Research Solution Selection Matrix that covers the Industrial Application Platform providers will assess the following vendors:

• • • ABB (Ability Genix)

    • Emerson/AspenTech

    • AVEVA CONNECT

    •  C3.ai

    • Honeywell (Forge for Industrial)

    • IFS.ai (Falkonry)

    • Oracle (Cloud Manufacturing)

    • PTC (ThingWorx)

    • SAP (Digital Manufacturing)

    • Siemens (Xcelerator)

    • SparkCognition

    • Symphony.ai (Manufacturing analytics)

Summary & Recommendations:

Over the past few years, Advanced Industrial Analytics (AIA) has quickly become one of the most overcrowded, diverse, and rapidly expanding industrial technology markets. Vendors in multiple shapes and forms serve this market, ranging from well-established Independent Software Vendors (ISVs) and Enterprise Resource Planning (ERP) vendors to large automation firms, numerous startups, and even major cloud infrastructure providers. As a result, it comes as no surprise that manufacturers today have an assortment of choices to choose from.

This blog post (and the analogy to dining options) is intended to help manufacturers simplify this complex process and highlight the choices companies have to make as they navigate this complicated vendor landscape. Future blog posts on this topic will dive deeper into the individual vendors listed in each category. Meanwhile, here are a few recommendations to keep in mind while selecting Advanced Industrial Analytics solutions:

• • • Don’t force this analogy on your software selection process: As mentioned above, even the best analogies only work to a certain extent; I’m aware this one has flaws, too. The purpose of this analogy is only to help visualize the vendor landscape in a different dimension and not to drive your entire software selection approach, which should be based on your business objectives, vendor commitment threshold, IT involvement, time-to-value, etc.

    • Start with user personas and then map use cases: The most common way to apply AIA is by beginning with the use cases and identifying business problems. While that is not a bad approach, beginning only with the use case without considering how it impacts different user personas creates organizational imbalance and inhibits scaling the initiative over time. Instead, start by identifying user personas that will be leveraging AIA, what their individual goals, objectives, and priorities are, and then map use cases.

    • The three approaches mentioned above are not mutually exclusive: Most medium to large-size manufacturers will likely need more than one option. Some might need a data platform approach and a few Best-of-Breed applications; others might have built their data platform in-house and want to leverage just the applications from a data platform provider. Either way, manufacturers need to consider which method they are following and take appropriate actions to advance their Advanced Industrial Analytics solution selection.