Your browser is not supported. Please update it.

30 June 2017

GE launches Avitas Systems in AI-driven robotics expansion

GE Ventures has launched Avitas Systems, a new division that will focus on using robotics, AI-based technologies, and predictive data analytics, to provide advanced inspection services for the oil and gas, transportation, and energy industries. The news was announced the day after long-time CEO Jeff Immelt announced that he would be stepping down, after leading GE through a period that has seen the company refocus on IIoT services – shedding divisions that didn’t suit this vision.

Launched at GE’s Minds + Machines conference, Avitas’ goal is to improve the efficiency of the kinds of mandatory operational inspections required in industrial facilities – especially those in remote locations. There are big cost benefits to be had from using something like a drone to patrol a pipeline instead of using a human worker, and in dangerous environments, there are also considerable safety improvements available.

Avitas claims that savings of up to 25% are possible, thanks to the ability to perform inspections based on ‘anticipated risk,’ rather than regular time tables – i.e. not wasting resources on an inspection that you are pretty sure won’t turn up anything new, but ordering an inspection if an operational data stream begins spitting out numbers that are not anomalous indicative of a fault.

Electricity distribution infrastructure, power plant chimneys, and cellular network towers, are all examples of equipment that could be more safely monitored by robots. Transport infrastructure like bridges or road surfaces might be more cost-effectively monitored by drones or autonomous vehicles, and of course, a key focus for Avitas will be the integration of emerging connected devices – to get access to all their valuable operational data.

Manual inspections are more likely to require manual data entry too. While the industry is embracing digitization, trying to adapt an established process to suit an IoT-based data-ingestion model is going to be rather tricky for developers. Greenfield opportunities, such as those where entirely new and sensor-rich equipment is being installed, should be much easier to automate – as the data-ingest should already be pretty seamless, thanks to the trend of cloud-based management.

The data collected can also be used to alter the inspection type or schedule, should a pattern emerge in the analysis that suggests a particular component or area requires more attention. That might be something that gets allocated to a dedicated human technician, but in many cases, it would be far cheaper to quickly dispatch a drone or autonomous vehicle to validate something in a data-stream, then to allocate the task to an organic worker.

Of course, the next step from inspection would be to resolve a problem once it has been identified. Should a machine or system report a problem, or an inspection discover one, an efficient system would set the wheels in motion to dispatch the appropriate technician with a truck full of the right tools to correct the problem.

In mission-critical environments, an inspection has the potential to trigger emergency shutdowns or repairs – meaning that they have to be certain of a problem’s existence, as that unplanned downtime can be extraordinarily expensive. Eventually, this kind of intervention could be automated by specialist machinery, capable of fixing problems with belts in production machinery, or performing diagnostic work as a human technician travels to the site.

Trials are already underway, with GE testing drones and crawlers in refineries, factories, and railroads. The company has said that the annual industrial spend on inspections is around $40bn, and using Avitas, it is hoping to carve a significant slice out of that market for itself.

“The inspection services industry requires cutting-edge technologies to help avoid unplanned asset downtime and to deliver new, valuable insights,” said Alex Tepper, Founder and Head of Corporate and Business Development at Avitas Systems. “We will use state-of-the-art robotics, automated defect recognition, and cloud-based technology to give customers the customized service and data they need to advance from reactive to predictive repair.”

GE isn’t the only company homing in on the IIoT. Close-rival Honeywell has been rather active, expanding its building and energy management portfolio, and its industrial automation offerings – which range from smart home, corporate office management, mining, oil & gas, manufacturing, and utility services.

IBM is also looking to its cloud platforms to expand into new industrial areas, using its Watson system and machine-learning to provide ‘cognitive’ computing services to industries looking to get more out of the data they already touch. The next step in that process will be to entwine itself more thoroughly with new devices, as they are deployed, in order to get its foot in the door of these potentially lucrative deals before they go live.

GE Ventures, as the name suggests, typically invests in startups, but is also responsible for launching new GE businesses. GE Current is probably the best-known example of an internal launch, which was set up to focus on lighting and energy, but GE Digital preceded Current, and was set up to house Predix.

The number of Avitas staff hasn’t been confirmed, but it will be based in Boston. The unit will be looking to build on GE’s expertise in the Avitas focus areas, as well its Predix platform – an application cloud that serves IoT apps, as well as handling predictive maintenance data analytics. This platform will let Avitas integrate data from external sources, such as weather data or CRM data, to add more color to the picture.