Learn the Keys to a Successful IoT Development Project in This Guide.

Subscribe to our blog to get the latest articles straight to your inbox.

For enterprises making the foray into IoT, using public cloud services for IoT deployments can decrease time-to-market and reduce upfront costs for on-premise data centers. However, it’s not the choice many enterprises today are making when it comes to managing their IoT devices. Instead, these businesses opt for privately owned and controlled servers. What are the reasons for this, and what does it mean for the IoT landscape?

Who Owns the Data in Enterprise IoT? 

Data is now one of the most valuable assets a company can access. It’s the key to improving the customer experience, creating new revenue streams, and improving operations with important insights.

However, there is currently no agreed-upon IoT data ownership model in the industry. That means that where and how you collect, store, and analyze your IoT data has huge implications for enterprises today.

Many companies are sitting on gold mines of data, and choices about the best uses for it are difficult. Data decisions range from the ethical (do my users know how I am using their data?), tactical (how can I use this data to get a competitive edge?), to purely financial (how much money is this data worth to someone else? Do I trust them?).

With an on-premise IoT solution, companies house and maintain the data in their own facilities, and if they choose can use their data centers to run private clouds as well. The public cloud doesn’t offer this level of visibility and control. Instead, it involves a third party in data decisions. 

In a panel discussion at the Internet of Manufacturing South conference in November 2019, Director of Advanced Analytics at Georgia-Pacific Ron Oudenaarden advised that the “Who owns the data?” question was a major factor in selecting a partner for IoT deployment. Georgia-Pacific believed that their company should own the IoT data they collect and generate, but many other development/platform partners they spoke with did not — and for them, it was a non-starter. 

Enterprise IoT Security and Regulations

If you’ve been keeping up with the latest news in IoT security, you know that in recent years,  the landscape has resembled the Wild West. It seems like hackers and researchers find new ways to exploit IoT networks every day, and it’s getting harder and harder to protect IoT devices as the attack surface grows.

Meanwhile, regulators continue to pass new laws governing data privacy. Breaking them can have expensive consequences, like Google’s $57 million fine for GDPR violation in France.

GDPR, in particular, has interesting implications for IoT as well, since the law requires companies to show the chain of custody for personally identifiable information (PII). This helps companies to prove to consumers and regulators that an individual’s information has actually been deleted under their “right to be forgotten.” Connected hardware and sensors introduce all sorts of new data points and transitions that make providing this audit trail more complex. 

It’s certainly possible to achieve highly secure data encryption in the cloud, and compliance with GDPR and future regulations can be accomplished with the right expertise. However, many enterprises have already developed compliance procedures and company firewalls to combat these issues, so choosing on-premise makes more sense because it fits with what they’ve already established.  

Efficiency, Disruption, and Latency in Enterprise IoT

While many companies today are using IoT in consumer-facing products, a large portion of businesses, like manufacturers, are using IoT to enhance performance in their critical infrastructure.

In these situations, where speed may be your main requirement, you should keep your data local — which makes on-premise a better option than the cloud, because decision-making happens at a local level in real time. Connectivity is higher between data centers and connected devices, meaning the lag is lower.   

Cloud services, meanwhile, perform data analysis in a centralized location not local to the connected device where the data is collected or generated. While this feature is helpful for computing data that will need to be shared with a wide audience, it could be adding unnecessary steps (and seconds) to the process if the data only matters for a small subset of users (e.g., one factory in a larger manufacturing operation).

Hybrid IoT solutions might also prove beneficial in these cases as well, although they come with their own challenges. Hybrid IoT often offers the optimal levels of control, speed, and the best location, but the solution is often complex and more difficult to maintain. 

Enterprise IoT Integration

Enterprises did not become enterprises overnight, so their existing systems and infrastructures aren’t brand-new, either.

Unless they’ve gone through or are embarking on an extensive digital transformation — an effort that requires massive amounts of time and resources, plus a shift in culture and mindset across the organization — many enterprises are working with legacy and/or custom systems installed decades ago before the cloud was the revolution that it is today. 

While integrating these systems with the cloud is possible for IoT, it can be more time and resource-intensive and could feel like trying to fit a square peg in a round hole. Making IoT on-premise or even hybrid may provide better integration and visibility. 

Your Next IoT Deployment

It’s hardly news anymore that cloud services offer enterprises speed, scalability, and operating cost savings.

That doesn’t mean, however, that everything will move to the cloud, or even that everything should move.

Product and engineering leaders have to address the realities in their own businesses and choose the option that makes sense for them — and that applies to IoT, too. What matters is that enterprises find IoT services partners that can be flexible and adapt to their needs, even as they grow and evolve. 

Looking for an IoT development partner? Let’s talk. Drop us a line today.