Top questions for CXOs starting IoT projects
"The Internet of Things (IoT) is coming and it's set to cause a revolution in workplaces everywhere. Global spending on IoT deployments will exceed $1 trillion by 2020, according to market intelligence firm the International Data Corporation. But before your organization commits to spending on IoT, here are the top questions C-level executives should be able to answer.
Is it scalable?
Taking an IoT project from the drawing board to final implementation is challenging even for the most experienced organizations. According to a survey by networking hardware firm Cisco, only 26 percent of companies consider their IoT initiatives to be successful, with 60 percent suffering from project delays. The most common cause of these unexpected setbacks is underestimating the complexity of scaling IoT deployments. The proliferation of cheap hard-drive capacity has resulted in a organizations developing a hoarding mentality, but the exponential rise of big data collected by IoT sensors means that this is no longer sustainable. It's crucial to define the scope of any IoT project in advance and to establish realistic expectations of how data is used, stored and deleted. In order to keep projects scalable it's important to perform as much processing and analytics as possible on the IoT devices themselves, and to avoid centralizing load on the network.
Is the right person leading?
Finding the right person to lead an IoT project is a crucial first step when planning a wide scale deployment. While it's not always essential for the project to be overseen by a C-level executive, typically the Chief Information Officer will have the most relevant experience for the job. In some cases it may be preferable to appoint a Chief IoT Officer who is specifically tasked with handling data infrastructure for smart device deployments. This should be someone with the prerequisite knowledge of digital technologies, network topology and information security. Having the right person in place to lead the project is key to organizational buy-in. If your organization lacks the in-house talent the job requires, it may be necessary reach out to third parties. More and more technology platforms are providing 'as-a-service' business models, which makes it easier than ever to partner with an experienced firm that knows how to deploy IoT.
Is it smart enough?
The IoT is increasingly converging with artificial intelligence (AI) innovations. According to technology research firm Gartner, more than 80 percent of enterprise IoT projects will incorporate AI by 2022. So how smart should your deployment be? There is no point implementing machine learning algorithms or neural networks for an IoT initiative unless it is absolutely necessary. AI will greatly increase the complexity and planning needed for any IT project so it should only be used where there is clear added value. That said, most IoT deployments will involve collecting data on an unprecedented scale, and that data is useless unless it is somehow transformed into insights. Often with IoT deployments there is no realistic way of handling the sheer scale of information that is collected using conventional techniques. If your project requires in-depth or real-time analytics, the chances are that machine learning techniques will be necessary to achieve project success. The important thing is to assess from the beginning how AI will be incorporated into the project and how you will manage the creep of growing complexity.
Is it secure?
One of the greatest challenges that organizations face with IoT is how to keep networks secure. By 2020, more than a quarter of all identified cyberattacks will target smart devices. The IoT encompasses an enormous array of devices from cameras and thermostats, to insulin pumps and the control systems for self-driving cars. Each of these computers has its own unique vulnerabilities and sensitivities, and a single unsecure component can leave entire networks open to attack. Securing IoT deployments involves the same risk management strategies that are employed for more traditional computer networks, but smart devices often involve more layers of risk and types of attack vector. As well as software security such as firewalls, anti-virus software and digital monitoring, IoT devices may need to be physically protected from external tampering. Currently, there is no universal standard or operating model for IoT security. This means that organizations must make difficult decisions about how to manage their level of exposure. In many instances, it may be safer to turn to cloud-based solutions that externalise as much of the digital infrastructure as possible, so that your team can prioritise securing the smart devices themselves.
Is it futureproof?
Deploying IoT across an organization can be a costly exercise and companies need to ensure that their networks will be interoperable with future developments. It can be difficult to anticipate all of the innovations that will impact digital infrastructure over the coming years, but there are ways of mitigating the harm of legacy systems. When defining how external sources will connect with your IoT infrastructure it's important to establish an appropriate Application Programming Interface (API). Be sure that the API uses open standards to avoid vendor lock-in and use software that is actively maintained by the open source community. When closed standards and proprietary software are used, there is no guarantee that they will be compatible with future data formats or will integrate with relevant online platforms. If you're able to answer all of the above questions then you're well on your way to taking your dream IoT project from theory and turning it into practice. There's never been a better time to take advantage of the declining costs of smart devices. With the right data at your fingertips, the IoT could transform your organization for the better."