The IoT Project Handbook

The Opportunity

The rate of IoT growth increases every day as test deployments scale quickly into extensive estates of devices harvesting data offering actionable insights for organizations. The Economist Intelligence Unit (EIU) Business index 2020 highlights this with the gradual drip of IoT devices into enterprises’ digital transformation programs turning into a tide of connected devices:

• 58% of companies are now deploying IoT technologies for internal use

• 45% of companies are using IoT as part of their external product and services strategy

• 57% of companies say they’ve seen more substantial returns on IoT investments than they expected.

The IoT Business Index 2020: A step change in adoption

 

Companies are capitalizing upon the insights found when integrating the data, sparking innovation, improving customer experiences and, reducing operating costs.

The Challenge

But there is a gap between the desire to change and realizing that change. For every successful deployment, there is an IoT project that never made it past approval and just as many that, while they reached deployment, proved ineffective. So, while 57% of businesses see their IoT estates delivering better-than-expected returns on investment (ROI)*, 43% of companies are still struggling to realize the value in their deployments.

IoT is Complex

Cost, complexity, and trust are all common concerns held by the many stakeholders, with new considerations that come to light as enterprises move from research, through proof of concept (POC) and into the ongoing lifecycle management of an IoT device estate. A leap of faith is required, and skills gaps bridged, to deliver an IoT solution useful across the entire enterprise’s ecosystem.

Mapping out digital transformation

IoT data, like any other data source, is a factor in an organization’s security posture, and as such, security is vital.  It doesn’t matter if the data source is a low-cost ultra-low-power sensor, its information may be integrated into a company’s information ecosystem, placing even the most modest device inside the machine room of an organization’s digital transformation strategy. The potential sophistication – think AI-enabled IoT devices – and the sheer quantity of devices means that a company’s IoT estate can quickly turn into an asset as valuable as any product or service. To protect that IoT asset and ensure it delivers ROI to its full potential, active monitoring is required. Therefore, it has never been more critical to proactively manage IoT device deployments to safeguard the trusted data they provide is secure.

About this Guide

IoT projects take time and effort. This guide seeks to help leaders and managers at each stage of an organization’s digital transformation.

 

To gain the most value, bookmark this guide in your browser and use it as an ongoing reference manual that helps:

  • You understand the drivers that will enable you to build consensus with key stakeholders
  • Convince decision-makers to invest resources in your IoT strategy
  • Highlight considerations at the specification stage
  • Offer a step-by-step guide to deployment
  • Manage your newly deployed device estate through its life cycle
  • Ensure you exceed your deployments expected ROI

Gaining Buy-In

 

 

 

 

 

 

 

Common Concerns, objections, and Drivers by Stakeholders

Understanding your Stakeholders

Convincing decision-makers to dedicate resources to your project is not always a simple task. Their drivers and concerns may be diverse, so it’s worth considering what they care about, be prepared for their objections, and use roadblocks as a learning exercise on the value and potential of IoT.

Click on the relevant stakeholders (Grey) in the chart below to build your business case:

Rearchitecting to Prepare for IoT

Forming a Project Team

One of the single most significant impediments to gaining buy-in and successful implementation of any digital transformation program can be the people involved, and existing structures and practices.  Projects typically entail deployments of connected sensors that improve internal efficiencies or form part of an external product. Regardless of whether the project is an internal or external IoT deployment, or if the primary beneficiary is a single department or the entire company, the chances are any IoT roll-out will likely need the support of several teams and a change in working practices.

Project team considerations list

Breaking Down Barriers to Unlock True Value

The EIU study found that 65% of respondents agreed that IoT was essential to their digital transformation plan, and cross-functional collaboration was vital in realizing ROI. The report also highlighted that companies needed to shift away from age-old assumptions and traditional departmental silos instead of focusing on the value IoT data presents.  With integrated IoT data and other data sets linked to cross-functional business processes and the broader value chain, companies amplify the insights gained. In short, the whole is greater than the sum of its parts.

 

But to realize this real value and trigger further IoT investment perpetuated by strong ROI, stakeholders need a paradigm shift. They must look beyond their departments and understand how multiple data sources within a company can offer far better insights to propel the entire company forward.

 

Internal Resource

Buy Versus Build

The odds of an organization having the resource to design a scalable and secure IoT solution inhouse are low. And even if that resource was available, the chances their people would have all the necessary skills in every area are even more remote. Areas that must be covered by the right expertise include:

The understandable skills gap that most organizations have at the start of an IoT project can be mitigated as they progress from proof-of-concept to extensive deployments – security is where we see this happening. However, most companies tend to look at achieving a minimum viable product rather than a fully-featured solution managed and monitored from the cradle to the grave. As a result, most companies will at least partially rely on external services and consultancy for the lifetime of their IoT deployments.

There are multiple key areas to address with early-stage projects to avoid unforeseen risks, additional costs, and distractions that delay a project:

Available security resources is a crucial consideration when evaluating your approach. The EIU report highlights that the vast majority of companies rely upon third-parties for security support in their initiatives’ early stages until a clear ROI justifies specialist security headcount.

Companies become more self-reliant in security as they advance with IoT

There are many considerations to factor in when embarking on a digital transformation, but weighting buy versus build is key. Unless you already have a large, under-utilized team in house with experience of executing IoT strategies, it’s always better to work with a trusted partner. A partner will reduce your time to market, mitigate risk and cost while allowing you to focus on value-add features.

Which of the following best describes your company’s approach to IoT security? (% by the highest level of IoT adoption)

What Does Success Look Like to you?

Metrics are crucial to understanding success during a project and the effectiveness of the deployed solution.  Due to IoT’s complex nature, there is no predefined set of key performance indicators (KPIs) defining the success or failure. Metrics depend upon the components that make up a device estate, where you are in your project, and your overarching objectives. But some typical parameters could include:

The number of KPIs will grow as you move from proof of concept to extensive deployment, so it makes sense to identify and monitor relevant KPIs as early as possible in the project to ensure time to market is reduced and ROI is evident.

Specifying your Device Estate

Considerations at Specification Stage

The specification stage is not just about identifying the device’s components or feature-set at the point of connection or provisioning in the factory. You may want to consider a whole range of factors to ensure your deployment remains connected, secure, and manageable for its working life.

It would be best if you were considering the device type and the tasks they perform, which dictates your connectivity options, power consumption, and the Operating System [OS] that they run. For instance, an ultra-constrained device will spend most of its operational life asleep and only have enough power to relay a kilobyte of data to nearby devices. In contrast, a mains or battery-powered resource-rich node will utilize cellular or Wi-Fi and comprise multiple CPUs capable of sending gigabytes of data every second.

There’s no ‘one size fits all’ approach and various device types often operate in a single network, utilizing a range of connectivity technologies. Connectivity considerations are a careful balance between range and bandwidth, coming at the expense of battery life.

Devices that send small packets of bidirectional data will add further complexity to your device estate, especially when data bursts are short and intermittent. It may be worth considering a gateway setup that can handle IP and non-IP data, runs a fit-for-purpose OS, and can save energy by reducing the burden on constrained nodes.

Having considered the devices, their connectivity, and how they mesh together, it’s time to specify a suitable IoT OS. Memory availability also varies significantly between device types. Hence, it’s essential to choose a device with flash memory and an OS that reflects your intended use.

The OS provides an interface between software and hardware that utilizes APIs to relay between the various components and the data management platform, no matter how diverse the device type. Most operating systems are flexible and will work with any cloud solution, but the cost, support, and time spent implementing and integrating is crucial.

Integration can take up the vast majority of a project’s resources, with device stacks consuming up to 50% of development time. Another 30% is taken up by connecting an application to a device management service.  Some organizations prefer the customization, power efficiency, and control that a bare metal instance offers, but fail to realize the risk, time, and cost associated with developing and updating this approach. Conversely, at the opposite end of the scale, the feature-rich Linux OS offers distinct benefits; featured in the table below:

Features by OS

OS composition

Head to Arm’s video on demand for more information:

See More

But all these benefits come at a price and place demands upon memory and the processor. Some device estates chose a real-time operating system [RTOS] like Mbed that offers some of Linux’s convenience, but without the associated constraints and the ability to safely stop devices (like a robotic arm) in real-time.

To Cloud or Not to Cloud?

Where will you host your device management service? Using a cloud-based software-as-a-service (SaaS) product offering is undoubtedly an obvious option and can offer a low cost-of-entry and a quick-start solution. Additionally, the provider carries out maintenance, and scaling up (or down) is frictionless.

A Comparison of Cloud and On-Prem Device Management Hosting

However, an on-premises alternative provides additional control and customization. It might be a better choice for those organizations facing challenges such as regulatory compliance, privacy, security, or technical oversight. Also, a public on-prem option facilitates integrations with other platforms, making it ideal for any organization looking to develop a white-label IoT solution for third parties.

 

 

Which Deployment Option is Right for me?

 

Further Reading

KEPCO Provisions Smart Meters to 30 Million Homes Case Study

Read More

Secure On-Premise Management of DistributedIoT Device Deployments White paper

Read More

You may want to deploy a turnkey solution to a customer data center with relative ease. Regardless of deployment options, it offers a ‘rinse-and-repeat’ approach to new business models such as infrastructure as a Service (IaaS) by deploying identical features and capabilities with ease.

Security is not a Consideration – It’s a Foundation

You may be asking why the matter of security did not feature, and quite prominently, in our considerations table. That’s because devices need to be secure by design, not just as a stage in the specification process. Security must be a common thread that flows throughout.

The rise of deployments creates a more significant security risk, and with every device deployed, the attack surface for potential hackers to exploit also grows. Additionally, the diverse nature of IoT hardware and their varied use-cases only adds complexity as managing a resource-rich gateway node on a factory line is significantly different from protecting a constrained fitness tracker on a jogger’s wrist. Furthermore, a device’s threat model and the methods for maintaining an acceptable posture can evolve over the life cycle.

Despite all these threats, there is no independent security testing available to IoT devices, which creates complexity for security framework providers, resulting in slower adoption of IoT devices. Insights that facilitate the business-critical decisions must have a basis in trust. Trust begins in the factory with a trusted chip that enables the secure collection of data delivered securely to a trusted enterprise ingestion platform that facilitates secure cross-application collaboration.

Further Reading

Securing IoT by Design: A Cyber Security White Paper by Copper Horse

Read More

PSA Certified

Read More

Confusion over what constitutes a secure device undermines trust, and a lack of trust impedes the adoption of IoT. However, the introduction of the Platform Security Architecture (PSA) can help developers with accreditation via a set of design principles without significant resources or skills. PSA helps highlight the security needs as it moves from design to factory provisioning and updates and is then eventually decommissioned.

Say Hello to the Edge

Just over a quarter of survey respondents to the EIU survey (26%) say that IoT data is pivotal to their current or planned use of Artificial Intelligence (AI), with 56% identifying IoT as “one of many important sources” for AI initiatives. Combining those two stats shows that 82% of enterprises see AI as key to realizing IoT data’s value and that IoT powers the growth in AI for business applications.

Computing that facilitates data processing at the edge of the network has emerged as a dominant industry trend and has caused the balance between the Edge and the Cloud to shift. As IoT deployments become more pervasive and expansive, a new generation of gateways is uniting legacy and IP devices and facilitating further advances. These gateways reduce bandwidth, improve latency, enhance security and privacy, and pave the way for applications that incorporate machine learning and artificial intelligence.

Should I Head to the Edge?

If you answer is ‘yes’ to one or more of the following questions; it’s worth considering placing some of your computational processes at the edge of your network:

Gateway specification – the pertinent questions:

 

Further Reading

Securing IoT by Design: A Cyber Security White Paper by Copper Horse

Read More
  • How will I collect data and optimize data?
  • What protocols are used, and how will they be translated?
  • How will I discover devices?
  • How will the message bus enable local command and control between individual components?
  • How will I deploy new analytics and functionality at the edge on an ongoing basis?
  • How will I update firmware and certificates securely?

Managing an Ecosystem’s Device Life Cycle

Ongoing Management

The specification stage is not just about identifying the device’s components or feature-set at the point of connection or provisioning in the factory. You may want to consider a whole range of factors to ensure your deployment remains connected, secure, and manageable for its working life.

Further Reading

Creating a secure device life cycle (compendium)

Read More

Turning Data into Insights

A deployment is born out a need and serves as the north star for your project. While project objectives vary, they all serve as a useful endpoint from which you can work back.

Data can:
• Spark new waves of innovation
• Unlock new revenue opportunities
• Reduce cost
• Inspire new working practices and processes
• Create new business models or markets
• Helped inform strategic or tactical decision making

Having a clear picture of what data can do for you will define all the considerations you make along your journey and ensure you do not encounter scope creep.

Is your Data Being used to its Full Potential?

Companies make progress with IoT deployments by focusing on IoT data’s value, applying it to one clearly defined business application that demonstrates a return on investment before growing into additional data sets. They then introduce advanced analytics and applying to cross-organizational value chains. When additional data sets complement that of a sensor, the insights gained are not four times more valuable. No, the value is exponential because decision-making tools available become far more valuable by offering more accurate data that combines more datasets to provide better insights.

Yet, conversely to break down organizational barriers and get past an initial proof of concept and involve cross-functional datasets requires a leap of faith, based on no data- just a strong business case with a clear end goal. But by identifying a clear purpose and addressing the considerations in this guide, you will be able to influence stakeholders, break down silos, and win the IoT.

Choosing a Partner

A device management partner can help guide you through your IoT journey. From gaining buy-in, by specifying a secure, scalable device estate that delivers ROI, abstract complexity from the estate’s ongoing management with a device management service.

A potential device management partner should be asking insightful, probing, and relevant questions regarding your device management requirements and your overall digital transformation strategy. Ensure the vendor you’re considering to be your partner has discussed the issues detailed here part of a consultative approach. The questions you ask at this stage could make the difference between a solution that supports only some of your requirements today or one that fulfills your future IoT aspirations.

Head to our on-demand seminars to view these topics discussed in greater depth:

https://www.arm.com/company/events/on-demand-webinars/iot-devices-vseminar

About Pelion Device Management

Pelion Device Management offers secure, future-proof, and turnkey IoT life cycle management from the device experts, delivering remote management of any IoT device from anywhere – cloud, on-premise, edge.

*Citation: Economist Intelligence Unit IoT Barometer 2020