7 Common IoT Mistakes and How to Avoid Them
Learn from other's mistakes.
Join the DZone community and get the full member experience.Join For Free
“ How can IoT work in my business?”
This question is the first thing many people think when they hear about all of the exciting possibilities in the IoT space.
But often, from there, the excuses start pouring in:
“My organization is too small to adopt new technology.”
“I’m too busy to do this now.”
“I work at a company that is so large, it is hard to figure out who should lead this project.”
Or the big one: “I don’t have the budget for this kind of project.”
The truth is that any business can now find some way to implement connected sensors and IoT technologies.
The success of a given IoT project, however, is not always guaranteed.
Too often, a failed pilot can be a barrier to other implementations, so it’s important to avoid making IoT mistakes that could cost you executive buy-in for future applications.
A Tried-and-True Recipe for IoT Success
Over the past few years at Temboo, we’ve worked with Fortune 500 companies, startups, cities, research teams, and local businesses helping them make their products, processes, and facilities ‘smart.’
Despite all their differences, the most successful projects hit a certain sweet spot in putting the Internet of Things to work for them and managed to either avoid or overcome some fairly common IoT mistakes along the way.
In this post, we’ll go over the biggest IoT mistakes we’ve seen clients make in their projects and how the most successful applications avoided them.
The Top 7 IoT Mistakes to Avoid
1. There's No One Problem to Solve or Too Many Problems
A few months back, a major telco reached out with the goal of developing a managed enterprise product in IoT, with a focus on predictive maintenance.
They had about 250 detailed questions for us before they could even start talking about the project. And yet, the company itself was unable to answer basic questions, like “What type of problems or equipment do you want to use predictive maintenance for?”
They had a slide deck, a detailed timeline, and an experienced team, but the company didn’t have a problem or even a problem area that they were focusing on.
Needless to say, their schedule stretched out and their deadlines were pushed further and further out past the horizon.
On the other end of the spectrum, we once spoke with a building manager that wanted a smart building app for their complex.
Not only did it need to be available as a smartphone app, but it also needed to be accessible via a website and have the ability to control public information screens throughout the building site.
Tenants, visitors, maintenance staff, event planners, and administrators all needed to be able to use the app in different ways, like publicizing events, internal messaging, tracking expenses, updating maintenance issues, and more. The app would be for everyone and do pretty much everything. We had to pass on that one.
The IoT Mistake: It may be easy to think you need to ‘get into the IoT space’ or that IoT can solve all your problems. But this kind of thinking is itself a problem and will reduce your chances of success.
How To Avoid It: Instead, try focusing on one clearly defined problem at first. And ideally, it will involve just one type of machine or one type of data or one type of activity. It’s smart to start simple.
2. The Scope Is Undefined or All-Encompassing
Even if you’ve managed to narrow down your problem area, the scope of your project can still trip you up. A tooling company we worked with faced this issue.
They had multiple machines from different vendors and of different ages that shaped metal into their various products.
The cold water system for their machines was their target area. This water cools the bits that grind down the metal, ensuring that it comes out shaped correctly into the nuts, bolts, and fasteners being produced.
The problem was clearly defined: If the water isn’t the right temperature, the final products aren’t shaped correctly and have to be scrapped. Better temperature monitoring of the water should help prevent the problem and reduce waste.
But how far should this first project go?
The tooling company at first wasn’t sure how many machines this should be tried on, when they’d be able to start, what success would look like, or whether they were any restrictions on the technology they could use.
They were going to need some help scoping this out.
Then, another colleague at the company jumped on the project and suddenly the scope expanded into baroque detail. All 25 tooling machines needed to be covered, nothing could be wireless, no cloud could be involved, any additional hardware for sensing had to unnecessarily include PLCs since that is what they were familiar with, and a budget was set in stone.
The IoT Mistake: When implementing new technologies, it can be tempting to be vague since there are many unknowns, but that’s no excuse for not scoping out the key elements of what the project should include.
How To Avoid It: When implementing IoT to solve a problem, the scope should contain a range of possibilities for how to achieve the desired outcome.
3. Everything Is Contracted Out or Is Being Reinvented In-House
When it comes to IoT, companies are often tempted to contract everything out to consultants and third-party development teams since they’re unlikely to have in-house experience or talent.
I saw this last year with a major appliance company.
They wanted to collect diagnostic data from their products to help service their warranty commitments, and so, they had hired one of the big three consulting firms and a team of outside developers to implement this.
This project would ultimately involve millions of their products and be driving a big portion of new revenue to their business, yet they still had only a handful of people in-house who could even keep track of what their consultants and developers were building.
It was hard to see how the company would be able to capture a lot of the gains of this new product development if they weren’t going to be able to build up knowledge in-house for the next IoT project.
But it’s also easy to make a mistake in the opposite direction and insist that everything must be built in-house.
This problem plagued General Electric when they first started building Predix, their Industrial IoT software platform.
They went so far as to build their own cloud computing technology and infrastructure for the system to run on, which cost an enormous amount of money and slowed them down from addressing the actual problems their product would fix.
Ultimately, GE had to make the painful decision to change course and ended up building their platform on top of one of the existing commercially available cloud computing infrastructures.
The IoT Mistake: Not building up in house knowledge of your IoT implementation or trying to reinvent the wheel with your application.
How to Avoid It: The key here is to find a middle ground where your company is getting its hands dirty developing some of the technology while also making use of the ever-expanding of the ecosystem of existing IoT hardware and software platforms. This way, your organization can build up competencies in-house but still move fast enough.
4. There’s No Single Leader or Everyone’s a Decision-Maker
Several years ago, we partnered with a major semiconductor company to add IoT capabilities for one of its new hardware platforms.
It had taken months of working through various teams and people at the company to find the right one to work with.
That was the first problem.
The next problem was that the Internet of Things didn’t fall under any particular person or department’s role at the business. The org chart wasn’t any help to us or even to the people within the company when determining where to go next.
Eventually, we found a team and a product that was a fit.
We scoped out the project, set goals, and timelines, and agreed upon functionalities to deliver. Then, we had a working prototype.
It seemed like the next step would be some final testing and then agreeing upon a public release and marketing plan.
But in the process of rolling this out, each additional person brought onto the project apparently had veto power over items previously agreed upon and even fully built out.
Each time this involved re-explaining the entire project to new people who had increasingly less knowledge and less at stake in getting this released. Unfortunately, these hurdles ultimately proved insurmountable.
The IoT Mistake: Even technology companies can’t rely on their existing org charts and decision-making processes to make IoT a success.
How to Avoid It: In this case, I saw first-hand how IoT projects need a clear and dedicated decision-maker. Making the decision on who in the organization will own a given initiative is crucial to its success.
5. The Timeline Is Too Slow or Too Fast
Some businesses think ‘agile’ can mean ‘next week.’
A lighting company I worked with wanted a remote-controlled smart lighting demo for their sales event in less than a month.
They knew what they wanted to achieve and had a realistic sense of what the demo should be able to do. But their timeline was setting themselves up for failure, especially since this demo wasn’t just going to be shown once but multiple times over an event spanning a few days.
That required a level of consistency and robustness from the demo that would be close to impossible to obtain.
On the other side is a government agency we worked with. Their internal processes meant that it took at least a year from selecting an IoT technology to implementing it in a pilot.
A year is a long time in technology, and this lag meant that newer technologies useful for the pilot emerged after the first solutions were selected.
They’d effectively be implementing systems that were already outdated for a pilot when there were already clearly better alternatives out there.
Fortunately, this agency understood the problems its traditional procurement process were causing and created a more flexible process for its technology pilots going forward.
The IoT Mistake: Not having a realistic timeline for your IoT implementation.
How to Avoid It: Get a clear picture of what you’re building and speak with the people who will be directly involved. Once you have an idea of what an attainable timeline is, you’ll want to get started right away to gain a quick win.
6. Maintenance Is Underthought or Future Developments Are Overplanned
Once a product, process, or facility has been made ‘smart’ using the Internet of Things, who will make sure it stays smart?
Just like website and apps need to be updated to handle new security concerns, features, and operating systems, so too do IoT systems.
Companies like the now-defunct Quirky (with its Wink hub) and the still-dominant Facebook (with their Oculus headsets) have both left their customers with bricked devices after failing to ensure that their TLS certificates were renewed on time.
Clearly, there was a lack of planning about how these devices would be supported in the future.
But on the flip side, I’ve also seen projects fail because so much time was spent planning how to address every possible challenge or hiccup that could occur post-deployment.
One business I worked with was developing a new sensor product to detect leaks in the basement, but they became too focused on the future of building a sensor platform and the potential partnerships and product opportunities that they wanted to set themselves up for. They moved too slow because they were never ready to take the leap.
The IoT Mistake: Not thinking about the future can lead to security flaws, bricked hardware, and other IoT fails. But too much planning can lead to projects never getting off the ground.
How to Avoid It: Ensure your system is set up for scaling, and for any software updates, you’ll need in the future (OTA update capabilities are essential for this!) But also, make sure you don’t let the planning take over the entire project to the point where you can’t get started at all.
7. No Tolerance for Failure or a License Not to Plan for Success
If failure is going to cost you your job, then you’re not going to volunteer for any project that has a high chance of failure.
Similarly, you’ll likely set goals that easier to meet such that success itself may become meaningless.
Alternatively, a license to experiment without any consequences can lead to interesting projects without any practical business value.
With IoT, companies often struggle to find that middle path which can lead to success.
IoT leaders need to implement a lot of different use cases to get the most payback, generally at least 15 with the average payback continuing to rise until the thirtieth use case has been implemented.
The IoT Mistake: The first IoT project can’t be the one shot to prove IoT works, nor can a company responsibly commit to ‘exploring’ IoT indefinitely.
How To Avoid It: Plan for quick wins that can show the value of IoT in your organizations while leaving room for scaling up and new ideas.
Hitting the Sweet Spot
Avoiding these seven IoT mistakes requires taking a middle path away from the extremes of overcorrecting in any one direction.
The most successful projects we see in IoT involve companies with a clear but solvable problem, a flexible but clear scope, a desire to develop their own competencies without reinventing the wheel, a timeline that takes into account how long success will take, defined responsibilities for post-deployment, and a sense of the larger vision for the company’s IoT strategy.
Published at DZone with permission of Jessica Califano. See the original article here.
Opinions expressed by DZone contributors are their own.
MLOps: Definition, Importance, and Implementation
Multi-Stream Joins With SQL
DevOps Pipeline and Its Essential Tools
The Native Way To Configure Path Aliases in Frontend Projects