The transition to Industry 4.0 also has a major impact on maintenance. The keywords here are “smart maintenance” and the somewhat more specialized “predictive maintenance.” To ensure maintenance is as efficient and farsighted as possible, teams of scientists are in the process of developing and testing a variety of technologies. Even in the field of robotic applications, companies are working at full speed. But what exactly do the latest applications do and in which industries does it make sense to use them? In our series about innovative start-up solution for industry 4.0 we introduce 5 to watch.
Using robots in plant maintenance is nothing new – they’ve already been used in nuclear reactors for decades. Since that time, robots have continued to develop and improve. The problem: for a long time, robot manufacturers could almost only optimize the hardware, while the software continued to fall behind in direct comparison. But this deficit could soon be a thing of the past: the revolutionary magic word is “data processing algorithm,” which promises a new generation of robotic applications: modern hardware combined with integrated software solution.
The hardware (sensors) is usually responsible for data collection, while the software processes the data acquired. Applications just like these could one day be very valuable in the areas of process engineering and process manufacturing, especially in the case of connected machines. Through the optimized interplay between hardware and software, robots can immediately process and implement data. In concrete terms, that means, for example, that the wear and tear of machine parts can be calculated and predicted using sensor technology, real-time data, and extensive comparative values.
The benefits are clear: servicing no longer has to occur at predetermined, inflexible intervals, but can be carried out cost-effectively based on the current situation. The safety hazard is also reduced, as inspections in dangerous and hard-to-reach areas can be carried out by robots and/or drones, reducing the danger for employees. Opportunities like these could improve maintenance over the long term and thus play a key role in reducing costs, increasing plant availability, and minimizing failure rates. We take a look at five start-ups and their robots in greater detail below.
COMPANY: ANYBOTICS, ZURICH
Overview: ANYmal is a four-legged robot with autonomous navigation that’s largely used for inspection work in the energy and process manufacturing industries. Its applications include maintenance and reliability reports in the context of machine monitoring as well as sensor reading. The robot can also be used for safety assignments, for example, in the area of power supply. ANYmal is suitable for older oil facilities, where shorter inspection intervals are necessary to minimize the increased risk of failure.
Its design and performance data (high payload, three hours of running time, speed of 1 meter/ second) make it possible to carry out urgent routine inspections for a relatively affordable price. More data can thus be collected than before with the same costs and incorporated into maintenance. Due to limitations in software integration, more complicated tasks are only possible to a degree at present. But ANYbotics is already working on a new version of the robot designed specifically for certain conditions and safety requirements in the oil and gas industry. As a result, it will also be possible to use ANYmal in areas prone to explosion (ATEX/IECEx), among others.
COMPANY: BOSTON DYNAMICS, WALTHAM (USA)
Overview: Similar in its design, construction, and performance data, the SpotMini boasts exceptional mobility, making it suitable for inspection work, too. Unlike its competitors, Boston Dynamics doesn’t focus on specific industries in its development. The Spot- Mini also has an articulated arm that allows it to, for instance, open doors and thus enter closed-off areas. It also has the ability to learn, allowing it to optimize its capabilities with each repetition. Thanks to its articulated arm and the integrated learning algorithm, the SpotMini may one day be able to carry out minor repair work in addition to inspections. Further developing the algorithm could, for example, make it possible to use the robot in reliability tests on machines or machine parts, which would make the SpotMini a shining example of Boston Dynamics’ claim: “We pride ourselves in building machines that both break boundaries and work in the real world.”
MANUFACTURER: CAMBRIDGE, MA (USA)
Overview: The Harvard Ambulatory Microrobot (HAMR) is much smaller than the two robots already presented. The developers themselves refer to it as “cockroach size.” And it’s precisely this small size that accommodates much smaller areas of application than other models. Developed by Harvard University in collaboration with Rolls Royce, the HAMR is also less susceptible to damage than its competitors thanks to its compact design – even when it falls from higher up. It also offers a high working speed and is flexible when it comes to overcoming obstacles. These characteristics allow the HAMR to also be used for inspection in machines and engines that you’d otherwise have to stop and, in some cases, dismantle. Because disassembly, which can be very time-consuming and expensive for many companies particularly in process manufacturing, is no longer necessary, the HAMR can play a key role in maintenance work that ultimately conserves resources. But the HAMR does not yet offer its own software solution – solutions would have to be purchased separately.
COMPANY: EELUME, TRONDHEIM, NORWAY
Overview: Much like ANYbotics, Eelume is a start-up originating from a university research institute. The construction of the Norwegian application is reminiscent of a snake. Eelume is suitable for underwater applications and used for inspections and repair work on offshore plants. After an extensive testing phase, Eelume received the necessary certification. The first pilot project will begin in summer 2019 with the inspection of Equinor’s underwater natural gas field, Åsgard. The snake-like body of the Eelume robot allows it to autonomously carry out more minor work at areas that would otherwise be difficult to access. Using an underwater docking station, the robot can be equipped with a variety of modules such as a brush head or gripper. The different modules and heads enable it to open and close valves as well as inspect underwater pipelines. While Eelume cannot yet analyze maintenance-relevant data, it will one day be equipped with artificial intelligence in order to further expand its performance in maintenance work.
MANUFACTURER: SKYSPECS, ANN ARBOR (USA)
Overview: SkySpecs offers inspection solutions for the wind energy sector, with a focus on maintaining wind turbines – thanks to the use of drone technology on and offshore. According to a company statement, the inspections “are completely automated, requiring only the push of a single button through to landing – and take less than 15 minutes.” SkySpecs has taken the opposite approach to the start-ups presented before: the hardware – in this case in the form of drones – is purchased externally, while SkySpecs has developed an analysis software itself. The drones are used to maintain wind turbines and can detect damage. The data collected are transmitted directly to the analysis software and analyzed. Benefits for the employees: they can create maintenance reports and other reports more quickly and, if necessary, make last-minute decisions about any repairs. While employees have to analyze the transmitted data at present, SkySpecs is working on an algorithm that one day will be able to analyze data directly. Although a majority of the wind energy sector still relies on manual inspection when it comes to turbine maintenance, the software solution could make SkySpecs inspection much more efficient and less risky.
Summary: Maintenance Sector Is on the Brink of Accelerated Development
The presented applications reveal the direction in which maintenance is moving as a result of technologization. Most exciting will be the areas in which robots combine with intelligent software solutions, bringing us ever closer to the transition from mere data acquisition to the creation of maintenance reports with concrete instructions. In a scenario like this, the robot would collect data and then transmit it to an ERP software (e.g. SAP PM), which would then plan any required or future repairs.
An alternative option for developing smart maintenance reports would be to provide robot start-ups with manufacturer data directly. This direct method would offer the benefit that robots would have data at their disposal from the start, rather than first having to collect them. In addition to manufacturer data, you could also incorporate the experience of industry experts when it comes to defining the maintenance strategy’s areas of focus. Using advanced data analysis software and artificial intelligence, robots would even be able to autonomously collect and condense maintenance-relevant data in a targeted fashion, which would further increase the feasibility of today’s “preventive maintenance strategy.”