No bed (or dog bed) manufacturer would dream of putting their bed frames and mattresses on sale without undertaking a prolonged stage of stress testing often using robots, rigs and jigs to simulate real-world wear and tear – so why should IoT kit be any different? Answer: it shouldn’t.
Internet of Business recently toured the facilities at the European Strategic Energy Technologies Information System (SETIS) in the Dutch town of Petten on the outskirts of Amsterdam. This is a secure site working on areas such as new-age fuel cell stress testing.
It’s not enough to just build renewable fuels and bioenergy-empowered smart fuel cells; we also need to know how long they are going to last.
Further still, we need to know how long they will last when deployed, for example, in a normal car that travels on smooth roads in comparison to the wear and tear experienced if the same unit were in stalled in, let’s say, an agricultural tractor.
What about fuel cells in humid countries? What about fuel cells in heavyweight industrial applications? We need to stress test the IoT and SETIS has the environment to simulate every condition imaginable.
Memory stress
Embedded memory products company Micron agrees. The company says that designers for the industrial IoT must make careful choices to ensure their products meet demanding performance requirements.
Micron’s new embedded USB (eU500) is a ruggedized storage device that enables very fast boot-up times and data-logging capabilities for demanding environments.
For example, in the mobile industry, the eU500 helps enable mobile operators to maximize the uptime of base stations deployed in remote, unmanaged locations, where fast reboot times are needed to approach the goal of 99.999% network operating uptime.
Read more: Energy: How ENEL is using IoT to embrace the ‘energy revolution’
Strategic differentiators
According to Micron, “As data growth increases and solutions become more specialized, memory and storage solutions are becoming strategic differentiators for our customers and partners and making possible a range of breakthrough capabilities from autonomous driving to virtual reality.”
What we can see here is a problem that reflects the already well-documented IoT security issue. That is to say, the IoT was built thinking about connectivity, data and device form factors. It was not built with security and and the interconnectivity of application programming interface (API) architectures in mind.
Similarly, we have, by and large, built the IoT and even the industrial IoT (IIoT) without thinking about longer term stress, wear & tear and resiliency.
Read more: Researchers create energy-efficient power converter for IoT
Soft on the inside, crunchy on the outside
Quite apart from any of these essentially hardware-focused factors, the discipline of software testing for IoT production deployments is a whole subject in and of itself. Software testing will involve data load testing, application execution testing, code exception testing and dependency configuration testing.
In other (less technical) words, IoT software testing will involve pushing the application to ‘drink from a firehose’, in terms of data ingestion and/or the connections and calls made to it. When we can break it, then we know how strong it is.
Our RIST-IoT future
This entire situation is increasingly being addressed and firms like GE have established dedicated Proof of Concept (PoC) labs and ‘foundry’ locations to test products in situ.
But it’s still true to say that we had the IoT before we had the IIoT. The Ruggedized Industrial Stress Tested IoT (RISTIoT) may never yet be coined as a term, but we do need it to happen.
Image credit: Orvis
Read more: Energy: How Eneco transitioned from commodity provider to service provider with IoT