Verticals, the Uptake way

Readers know I have lamented for a while that enterprise software has paid lip service to verticals. Two decades of cloud and in-memory computing and still most of the available functionality is around horizontal areas - accounting, HCM and CRM. In the meantime, customers continue with client/server and older architectures around utility billing, retail merchandising, electronic health records, insurance claims processing, hospitality reservations and countless other “books of record”.  And even in verticals they have kept investing in, functionality has not kept up with changing trends. So, in manufacturing which many ERP vendors have supported for decades the functionality seems oblivious to growing digital capture via wearables on the shop floor, growing automation in the form of robotics and 3D printing and other contemporary practices.

So, I was a pig in mud this week visiting with Uptake in Chicago and hearing them talk about failure prediction, fuel management and label correction around complex industrial assets like locomotives and wind turbines in industries like railroads, mining and energy. These are industries with complex, expensive, mission-critical assets.  Connecting such assets is allowing for predictive maintenance, better asset performance monitoring and much more sophisticated field service.

Uptake is the brainchild of Brad Keywell and fellow serial entrepreneurs who have transformed Chicago’s tech landscape – see my post here.

The concept of having assets talk to us about how they are feeling is not new. In The New Polymath I described how as far back as 2004, BP was deploying motes, sensors, RFID tags in its tankers and in refineries in what it called a “sensory network”. In later books and blogs, I have extensively covered GE and its vision of the Industrial Internet.  Over the last couple of decades, our knowledge of complex assets, their components and failure rates and root causes of failures has exploded. Sophisticated companies now routinely make decisions about trade off between cost of unplanned maintenance versus cost of downtime.

Enterprise software vendors should have been all over this opportunity but most still just glibly talk about the “internet of things”. Uptake’s platform built atop cloud infrastructure from Google, Amazon and Alibaba ingests gobs of asset sensor data, applies machine learning to separate signal from noise, supplements traditional physical models and leverages a growing army of of domain-aware data scientists. It has lined up a group of demanding customers – 50+ so far in its 4 year history. Its platform can increasingly support a wide range of industries as it can apply its libraries horizontally to an expanding portfolio of use cases.

Uptake will not magically make up for lack of updated transactional capability from software vendors. As an example, most asset management software in the market still focuses on depreciation accounting and planned, not predictive, maintenance. Some continue to call their modules “fixed assets” and their last innovation was reading bar codes to track asset location.  Uptake should allow customers to continue to leverage such older transactional software, move them to cheaper third party maintenance and other “coping strategies” I described in SAP Nation.

In that sense, Uptake is showing that the value is moving away from process and transactional capability to intelligence and analytics.  Enterprise software vendors will attempt to do the same or at least talk a good game. Which is why Uptake’s most recent executive recruit, its President Ganesh Bell, ex JD Edwards, PeopleSoft and SAP is a timely one.

Can Uptake go even further and adjust to the world of outcome based economics? OEMs like GE who make such assets are trying to move away from the business model of one-time asset sales to multi-year contracts based on uptime and other outcome metrics. Buyers are skittish about such value contracts because they cannot precisely tie payback to product functionality. Still others view ML algorithms as voodoo science. Then there is the growing battleground about who exactly owns the data and how best to secure it.

Uptake is today a small player surrounded by giant OEM makers and sophisticated utilities and railroad customers. Again here, Ganesh’s recent stint at GE Digital should help.

It’s investment in understanding complex assets and OT (as different from IT) should make it a key player for the next several years as we evolve field service (factoid – the Bureau of Labor Statistics says wind turbine techs were the fastest growing job in 2017) and move to outcome-based contracting.

In the meantime, through his portfolio of other companies, Brad and his colleagues have their fingers in consumer markets, logistics, healthcare, advertising and other sectors. When it comes to verticals  I have a feeling they have already contributed more and and we will see more from them than we have from many enterprise vendors.