Common causes of IIoT project failure and actionable framework for a successful IIoT implementation.
This article is written and contributed by Matt Isherwood, managing director at Pathfindr.
Matt Isherwood, Pathfindr
Manufacturing has recently been transformed by the evolution of Industrial Internet of Things (IIoT), with asset tracking a particular area where time saving gains can be made. Enabling connectivity within a historically analogue process, it can provide actionable intelligence that can either speed up production – by locating or tracking assets or, in some cases, enable organisations to completely rethink how they make things.
But, as is often the case with tech integration, you cannot simply buy an IIoT platform, implement it, and expect it to transform operations without first looking far more broadly at why and how you are doing it. By definition, the industry only hears about IIoT projects that are successful, so there’s a tendency to assume it always works, but in reality the vast majority of programmes fail. In the specific context of asset and process intelligence, this is usually either because of integration issues or because senior stakeholders within an organisation have not been convinced of their wider value beyond an initial trial or siloed implementation.
Through this article I’ll share some of the common causes of project failure, before providing an actionable framework for organisations looking to implement IIoT.
Asset and Process Intelligence Explained
Manufacturing and engineering professionals will be familiar with the concept of the ‘hidden factory’. It represents the untapped capacity within a manufacturing plant – the maximum amount of additional productivity that could be unlocked if only it was identified.
Many manufacturers experience huge difficulty and bottlenecks when trying to get the right parts to the right place at the right time. As a result, workers regularly have to down tools and wait for key assets and materials to arrive before they can move on to the next stage of the process.
We can illustrate the scale of this challenge. Pathfindr recently undertook a survey with more than 100 employees at one of the world’s leading engineering firms, and 99 per cent said that the loss of assets, even if only temporary, regularly affects their operations. When asked how long it takes on average to find an asset, responses ranged from 10 minutes to a few hours, a few days and in some cases, never.
This is the fundamental problem that IIoT has the potential to solve. Sensors can be placed on every tool, part and asset within a supply chain, creating huge operational visibility including the ability to locate assets, and use data to improve performance and reduce inefficiencies. This is asset intelligence.
The same sensors also enable manufacturers to map and understand in minute detail every single stage of a manufacturing process, to quickly and accurately note where efficiency gains can be made. This is process intelligence.
Where Things Go Wrong
When a manufacturing organisation explores the potential to implement IIoT, or asset intelligence specifically, it is when they have some idea that there is waste in the system. Some may have undertaken sophisticated analysis to quantify the extent of this, others may have a general sense or anecdotal evidence that inefficiencies exist and wish to limit them as far as possible.
Assuming there is a good case for adopting technology, one of the major stumbling blocks we see is around implementation, in particular software integration. Organisations now use multiple software platforms, often with overlapping functionality. When implementing IIoT, in many cases the optimal solution is to discontinue certain software platforms, but many are reluctant to do so give the initial investment. This can lead to fatal flaws in an IIoT project, either because performance and data accuracy are affected, or because data analytics can’t be performed centrally though one user-friendly platform.
This can lead to the second major stumbling block for an IIoT project – failure to demonstrate value quickly. In the vast majority of cases, new technology adoption begins with small trials, or as more established solutions with defined business units. For a technology solution to gain broader internal buy-in, it must demonstrate value, and quickly. If a solution doesn’t produce actionable data, or if tools aren’t used to their full capacity, a robust ROI case cannot be made.
Finally, technology is a major investment, and there has to be a reason for using it. Those that do create a more robust case for implementation are more likely to be effective. So what steps can an organisation take to identify and implement a solution?
The Asset Intelligence Ladder
The Asset Intelligence Ladder is a very simple organising principle accessible to businesses of all sizes and levels of digital sophistication and supports a focus on the opportunity (to reduce wasted time and improve process) rather than technology. It’s important not to let technology provide a limiting frame of reference initially.
Is therefore recommended to start the process with a blank canvas, focused exclusively on business operations and how an asset intelligence approach could drive efficiencies, improve processes and deliver better outcomes.
1. Agree on a project team
With that in mind, the first step is to agree on the people who are going to drive this forward. Within a small business, or a production cell, this may seem immediately obvious, but all organisations should consider including a range of business teams and job functions. This will be invaluable in understanding the real value of assets well as in exploring a range of scenarios.
2. Agree on ontology
The project team should address the key concepts and terms under discussion, including asset, attributes, intelligence and location. Formally described as an agreed ontology, this is vital to ensuring that all parties are talking the same language.
3. Classify and value assets
You should then classify assets against the agreed terms of reference – described as a taxonomy. Initially this should focus on financial value, ranking assets by their capital cost. It is vital to include the long tail in this initial audit, including raw materials, components, accessories, equipment and more. It is important to also consider assets’ second-order values, especially the business or operational impact of their unavailability or loss.
4. Explore operational scenarios
Now is the time to explore operational scenarios and their business implications in more detail. Many of the scenarios will have been outlined when addressing the second-order value of assets and, especially, the prevalence of asset wait time.
For example, it may be that Asset X is never proactively returned, or charged, after use by Cell Y, meaning that all subsequent users must first locate and then charge the asset before it can be redeployed effectively, creating multiple process inefficiencies along the way. No situation is too trivial to be considered – it is important that as many process frustrations and bottlenecks as possible are identified and shared.
5. Prioritize use cases
Having captured the widest range of scenarios it is then time to organise and prioritise these disparate elements into potential, prioritised use cases, and consider what a solution would need to achieve. Attributes are likely to start with location intelligence requirements e.g. indoor, outdoor or both, on-site, off-site or both, the level of geographic precision, and real time or not.
You may conclude that your requirements are so varied that multiple solutions are needed, supporting the Gartner forecast that by 2021, 65% of enterprises will be using three or more technologies to track hardware assets.
6. Create a business case
Ultimately, a business case will be determined by a comparison of the cost of waiting for assets, or the cost of process inefficiencies, versus the total-cost-of-ownership of the solution (including installation, maintenance, training etc.).
Assuming that you move ahead to a pilot project, it must be designed in such a way as to deliver a clear ROI to inform decisions on the scalability of the approach.
The intelligence gathered across each of these stages will be key to setting up for success by focusing on measurable business impact. What we see on a day-to-day basis working with manufacturers in aerospace, defence, energy and construction, reflects the broader trend towards widescale industrial adoption of asset and process intelligence. Not only are a broader range of industries implementing them, but they are starting to reach every corner of those organisations that conduct successful trials. The asset intelligence ladder principles will be vital in maintaining this trend.