The Industrial Internet of Things (IIoT) can pay big benefits for organizations that do it right. It’s no surprise that a report by LNS Research and MESA International shows that more than 50% of manufacturers plan to pursue IIoT in the next 12 months. In fact, one of our gas pipeline customers has already saved over $9.8 million in maintenance downtime costs primarily because of IIoT.

Avoiding downtime is a huge motivator for companies embarking on IIoT. Outdated operational technology (OT), such as physical sensors, proprietary control-system software, SCADA, and historians, represent significant downtime risk. When they fail, costs are high. According to the Aberdeen Group, the cost of downtime went up to $260,000 per hour on average between 2014 and 2016.

Here are four best practices about keeping your eyes on uptime as you embark on IIoT:

1. How much uptime is enough?

Downtime isn’t cheap. You already know the direct costs of delayed product deliveries, idle time and overtime pay, and repair expenditures. Your business may also need to factor in damage to reputation, environmental damage, litigation, and more. Stratus has a downtime calculator to get you started.

Once you know the hourly cost of an outage, determine your tolerance level. While 99% uptime may sound fantastic, that’s actually 88 hours of downtime in an average year—or potentially millions of dollars. Or would you prefer five minutes a year? That’s where fault-tolerant solutions pay dividends with 99.999% availability.

2. Protect before you recover

While a disaster recovery plan allows you to return to operation after a catastrophe, it’s not enough.

By the time disaster recovery kicks in, considerable damage is done—and often irretrievably. Data traveling in your production environment is primarily “inflight” data, occurring in milliseconds and requiring instantaneous response. That data is lost in an outage. Controllers, temperature monitors, and failure analytics can’t wait while systems resume operating.

You’ll want an availability solution that prevents outages from occurring, ensuring zero data loss.

3. Keep it simple

When merging IT and OT systems to achieve IIoT, simplicity is critical. Operational staff often doesn’t have the advanced IT skills needed to manage complex IT deployments. So choose an availability solution that is simple to deploy and easy to operate and manage. That way, you can focus on making sure your plants are performing well.

Virtualization is also essential to simplicity. Virtual machines are isolated from unexpected problems elsewhere, and servers can be smoothly migrated offline for orderly upgrades and updates. The challenge is that consolidating IT and OT systems on a single physical machine replaces multiple points of potential failure with a single point, increasing your risk exposure.

You can overcome this risk by deploying a hardware or software-based continuous availability system. Such solutions run on standard-type servers in virtualized environments, and requires no special expertise to maintain.

4. Solve today’s problems with an eye on the future

Most industrial automation organizations are not ready to perform a full-scale upgrade all at once. So phase your planning to solve real problems immediately. A couple of early wins will get your project off to a strong start.

Just leave ample room for future growth and modification. Build on industry standards and proven methodologies. An availability solution you can implement and forget about will pay itself back in future compatibility and productivity.

IIoT is no longer merely for the early adopter. When done right, IIoT can create enormous savings and competitive advantage. As you face this complex undertaking, remember to factor uptime into the equation. You’ll build a solid foundation for achieving genuine, measurable benefits.