Why not AI instead of an IoT Platform?

by | Jul 22, 2019

Many OEM’s are struggling with IoT. From my perspective, if you have more than 25% of your devices connected to the Internet and under management, you are doing quite well compared to everyone else.

In addition, OEM’s have come to an IoT crossroads today. Many are faced with the challenge of migrating from their current IoT platform, which they’ve been using for years, to another one. Others are earlier in their journey and are trying to make sense of the number and disparate quality of other IoT platforms that are out there.

The core team at RevTwo founded and grew http://www.axeda.com/, which we would argue was the first successful IoT platform. Its purpose was to enable OEM’s to proactively monitor their devices in the field, and if a device had a failure or an issue, someone would get an alert and then login to a dashboard and try and assess what was going on. It was continuous remote device management.

Early on, we identified the major KPI’s that would show that remote device management had value:

  • Reduced number of field service dispatches
  • Improved first-time fix rates
  • Reduced MTTR
  • Improved corrective maintenance/preventive maintenance ratio

However, it stopped there. What if the person in front of the device couldn’t triage and resolve the problem? Then it escalated to help desk, and if they couldn’t triage it then a field service engineer would be dispatched. If the device was connected to the Internet, then they could open up screen sharing or a remote access session to speed the process.

And what if the device wasn’t connected? An IoT platform wouldn’t help you there.

And the connectivity challenge is getting harder, not easier. Most everyone is familiar with increasing security concerns around connected devices. And it’s even harder in EMEA, with the GDPR laws enacted in 2018.

So if you’re at this crossroads, what do you do?

We started RevTwo AI with the goal of enabling companies to take that next step of being able to dramatically shorten triage time and problem resolution for complex products. It’s not continuous monitoring; it’s event-driven, being initiated when a support event occurs. We have shown that even though our AI is cloud-based, the target device doesn’t have to be connected to the Internet in order for the AI to work. We have shown that using AI can also show improvement in these KPI’s:

  • Reduced number of field service dispatches
  • Improved first-time fix rates
  • Reduced MTTR
  • Improved corrective maintenance/preventive maintenance ratio

Hmmm. Look familiar? Additional side benefits:

  • The device doesn’t have to be connected
  • No more log file “sneakernet”
  • Customers are equipped to resolve more issues before escalating to support

So if you’re at this crossroads, should you be looking at an IoT platform? Or should you be looking at AI? It improves the same KPI’s and avoids much of the connectivity challenges of IoT platforms. And it’s less expensive.

If you want to learn more, send me a note at dave@revtwo.com