Blogs

Why do IoT projects fail?

February 5, 2020

Posted by: Anasia D'mello

There was a time when the fact that large IT projects went way over budget and were delivered well past the delivery date wasn’t particularly unusual; it was almost par for the course. So why should the Internet of Things (IoT), which encompasses computing and communications technologies be any different, asks freelance technology writer, Bob Emmerson?

Moreover, large IoT projects are intrinsically complex for other reasons, so there must be failures. End users don’t talk about them, that would be a dumb move, but some vendors are. For example, in May 2017 Cisco was indicating a failure rate of 75% and in July 2019 Microsoft stated that 30% of IoT projects fail in the proof-of-concept stage.

A recent Beecham Research 100-page report details and quantifies the key reasons. They included business issues such as the need for a well-defined strategy and a clear set of objectives, as well as technical concerns about security and IoT’s complex mix of communications and computing technologies.

Interviewees were naturally more inclined to discuss success rather than failure.  Some 42% considered their projects to have been either fully successful (12%) or mostly successful (30%) while 58% considered their IoT projects to have been either not successful at all (18%) or mostly unsuccessful (40%). The report, “Why IoT Projects Fail”, is available here.

Brutal honesty

This is my take. Being brutally honest about failures is not the regular pitch of IoT vendors, which goes like this: The market is booming and by 2025, there will be more than 41 billion devices; IoT is a key component of the digital transformation process along with edge computing, 5G and AI, etc., etc. So, get on the train before it’s too late and success is assured if you partner with us.

Warning the market about failure rates, and then going on to indicate that in most cases technology wasn’t the problem, is refreshingly different. A vendor that is upfront with this information and then goes on to indicate that most problems arise from client company culture, organisation and structure is more likely to come across as a credible solution provider.

KORE cites lack of vision or strategy as being the biggest challenge and highlights the need to conduct an “IoT Readiness Assessment”. In a nutshell this means evaluating the company’s technical resources and using this as a baseline, for creating and executing an attainable IoT strategy. They also recommend employing IoT strategies that start small and which target specific business processes. Then move on to larger projects using that experience.

An interesting variant on this approach is to employ an off-the-shelf solution “kit” designed for a specific vertical and comprising the requisite hard- and software. This allows solutions to be deployed in a few days and to be customised at a later date. Libelium has a comprehensive range of kits; prices range from €3,000 to €8,000. Software AG markets a package that combines the company’s self-service Cumulocity IoT platform with pre-built solution accelerators and the services of experienced consultants.

However, while starting small, employing starter kits, and a service-service platform are attractive options, the resulting solution could fail if it paints the company into a silo corner.

Self-evident truths

Bob Emmerson

IoT is at a critical tipping point. On one hand, advances in technology mean that it’s cheaper and easier than ever to produce “smart” devices, equipment and machinery that deliver significant benefits; benefits that the business community cannot afford to ignore. On the other, there are equally significant challenges like skills shortages, security concerns, and solution complexity that may compromise the business benefits if not addressed.

There is no IoT silver bullet, but viable, cost-effective solutions are being realised by companies that partner with a vendor or solution provider whose ecosystem comprises the resources needed to create and deploy robust end-to-end systems. That is a big ask and offers should be carefully checked out.

IoT value chains are only as strong as the weakest link. Companies that have substantial IT resources can create their own ecosystem but an in-depth understanding of IoT Operational Technology (OT) is needed as well as the middleware needed to dissolve the barrier between the IT and OT domains. Microsoft, for example, has significant IT resources but it employs an ecosystem comprising more than 10,000 IoT and intelligent edge partners.

Focus on the big picture

The starting point for any IoT project should be the big picture, the strategy vision, what issues need to be addressed and what are the expectations. It is a mistake to try and do everything at once. Instead the project should be unpacked and if stand-alone component solutions are created, they need to avoid the silo trap and enable interoperability with other components, e.g. using APIs and open systems standards. And every decision taken during the project should link back to supporting that goal.

The same caveat applies to data. The IoT is predicated on the ability to acquire parameter and event data, but not everything that’s generated needs to be processed and analysed. Right now there is a data surfeit, companies already have more data than they can manage, therefore it is important to determine what data is needed in order to realise the strategy. Hitting a target is only useful if that target needs to be hit.

Big challenges: Equally big rewards

The business community continues to make significant investments in IoT. They can’t afford to stand still and companies need to innovate in order to stay competitive. However, now that the failure rates are out there on the ’Net, and awareness of the development pitfalls is growing, we should see a healthier, more realistic market in future.

The author is freelance technology writer, Bob Emmerson

Comment on this article below or via Twitter @IoTGN