The most important Industrial Tech exit to date – OSIsoft sells to AVEVA!
The exit of OSIsoft is a BIG DEAL in the industrial tech space. After 40 years, J. Patrick Kennedy finally sold his baby to AVEVA for $5B. I’m inclined to think this is a beacon for the space in general, indicating increasing maturity across the IIoT ecosystem. This is, quite simply, the smartest and best acquisition anyone has made in the space. To get to the heart of it, we need to first take a deeper a look… who exactly is OSIsoft?
You might call OSIsoft the first IIoT platform. Founded in 1980, OSIsoft developed the PI System, which is the dominant data historian in the market. A data historian is a system for capturing data from industrial operations, generally sensor and machine state/activity data, typically in a time series format. Until recently, most of this data sat dormant in facilities, only to be queried after events of interest happened. In recent years, historians have been used as foundations for more advanced analytics programs, and are increasingly used for real-time intelligence and situational awareness.
The amount of data captured in PI systems across the world is truly staggering – I’ll quote the press release here: “OSIsoft works with over 1,000 of the world's leading power and utilities companies, 38 of the Global Fortune Top 40 oil and gas companies, all of the Global Fortune Top 10 metals and mining companies, 37 out of 50 of the world's largest chemical and petrochemical companies, and 9 out of 10 of the Global Fortune Top 10 pharmaceutical companies.”
OSIsoft is everywhere. I’m normally a hype-buster, but it’s difficult to over-hype how prevalent OSIsoft is across industrial operations. While OSIsoft has managed to dominate in the historian category, the company has not had much success selling analytics tools that leverage the historian data. Instead, myriad vendors (literally every industrial analytics vendor) have built PI connectors for their own analytics offerings; meanwhile some formal partnerships have been developed to sell PI-based analytics solutions (PTC, SAP, MSFT, etc).
AVEVA is making a highly-strategic land-grab to become the de facto industrial software vendor that can directly leverage all the PI data. AVEVA has been selling various flavors of industrial software since 1967, and the company merged with Schneider Electric in 2018 – Schneider owns 60% of AVEVA. Like all the major industrial systems vendors, Schneider has been pushing IIoT/digital products for years, although none of the industrial systems vendors have figured out how to make a lot of money selling this stuff. The acquisition of OSIsoft might be just the thing to help Schneider emerge as the leader in this category.
Now the question is, what will AVEVA (and Schneider) do with OSIsoft? The last acquisition in the IIoT space that got me this excited was Cisco buying Jasper for $1.4B in 2016 – it was similar in that it was a land-grab for connectivity, and Cisco picked up a huge amount of connected machines in that transaction. But while lots of companies were excited about using the Jasper platform pre-2016, I have heard almost nothing about people using the Cisco Jasper platform since the acquisition (I’m inclined to think Cisco killed Jasper because it was competitive, as it promoted the use of mobile networks over Cisco’s Wi-Fi and fixed-line internet portfolio).
Will AVEVA get it right and leverage OSIsoft’s highly-strategic footprint to dominate the world of industrial data? I generally like to root for start-ups and new entrants, but I’m excited about the opportunity for AVEVA/Schneider to help the entire space leap out of the trough of disillusionment. The history of the industrial tech space is plagued with failures and a lack of overall growth/scale – GE should have bought OSIsoft and made it the foundation of the Predix platform… we’d be living in a different world if GE had gotten it right.
The only big companies really making money in this space are Microsoft and Amazon selling IIoT cloud solutions, but you could argue they’re just selling repackaged versions of their existing IaaS tools. You might also include Rockwell/PTC on the list, but I’m skeptical of them having any real revenues from selling IIoT solutions (just a whole lot of pilots). The acquisition of OSIsoft might be just the thing to help a big industrial vendor start winning real revenues in this space, and to show the world the true ROI potential of strong IIoT solutions.
By and large, industrial companies have been disillusioned with vendors in the digital industrial tech space. The big players have failed to deliver compelling solutions with the right business models – and while there are a handful of great start-ups, most large industrials are hesitant to go all-in with start-ups. We need a big (trusted, established) player to get it right and drive progress forward… so AVEVA/Schneider, we’re rooting for you!
PS – What does AVEVA do with all these OSIsoft partners who are in fact direct competitors? Can AVEVA simply cut these partners out? Probably not – too risky, would hurt all the companies that have built critical solutions on top of PI with these other players. It will be interesting to see this all play out…
UptimeAI is an artificial intelligence-based Operational Excellence (OEx) platform, that goes beyond predictive maintenance with explainable failure modes/recommendations, and self-learning workflow
4 年YOU ARE SPOT ON MR BROWN!!
Retired Senior Smart Manufacturing Advisor
4 年As one of the early users/implementers I was always exited with this solution! Started with 1 implementation and retired at +150 n one company! hope Aveva will maintain the heritage from Pat!
You were the first analyst to recognize pi as an iiot platform.
Available for short or long term projects in Instrumentation, PLC, Startup, and Pulp and Paper Consultations. Road ready! Looking for only contract with per diem work.
4 年Oh great, Schneider sucked up another good tech company!
Manufacturing IT, Laboratory , process automation systems & data Architect.
4 年Their software is 40 years old and it shows. Anyone that has ever installed it will be familiar with the myriad of separate windows services that must run.Not to mention the fact that you have to recreate the data model via asset framework every time for every control system. He sold in time. If future controllers come with OPCUA server and industry standard data model baked in, then pi is redundant.First expose the data via OPCUA server. An OPCUA client can subscribe to events and continuous data via pub sub, and store to no SQL cloud based DBs using the likes of Azure iot event hub . And you can do all that and it's easier than implementing pi with all it's different servers and duplicated data clients (interfaces). Once someone like open process automation forum prove out the reliability of this style of approach, there will be a non pi option . I can't wait.