ETSI pushes for standards in augmented reality

As lockdowns proceed around the world, the question for many industrial companies is how they organise themselves for what could be a lengthy period of travel restrictions that vary from month to month as infection rates change.

Many have found they could organise activities in the short-term by finding unexpected uses for social-media apps. It has not just been Zoom that has kept staff in contact. WhatsApp has turned out to be a handy way for people on location to ask: “Is this meant to go here?”

What made WhatsApp and Zoom so handy lay in their familiarity. People already knew how they work, for the most part, because they already used them socially. What some technology suppliers hope is that the management and staff will see the value of something that improves the sense of presence: augmented reality (AR).

The lockdown makes a solid case for AR. A specialist working from home does not have to engage in a conversation that resembles the 1970s UK TV show Golden Shot when guiding staff on location around a warehouse or a pipe assembly: “up, left, right a bit, that’s the one”. They can use a mouse or even data gloves to point out on the factory worker’s headset where to look and what to do next.

Mechanical CAD companies such as PTC have been promoting AR for some time, and it was among those that seized on the lockdown to promote its software. The company has been offering free access to the Vuforia Chalk application to potential users.

One issue that AR faces is a lack of interoperable standards. Many of the suppliers involved in AR at the moment are in the land-grab phase, where it makes more sense to try to control as much of the stack of software as they can. This is likely to be a problem for users over time as they try to reconcile the software environments they have with those that have the AR features they want. But if usage becomes more widespread, users are likely to demand interoperability rather than try to replace their CAD software or try to work with format converters, especially as it’s a complex space.

As the first step in an attempt to promote greater standardisation in AR, a working group at European standards body ETSI put together a reference framework. Published a couple of months ago, it shows how complicated the picture is. And it is not one that favours start-ups right now.

Muriel Deschanel, business development director at research institute B-Com and chair of ETSI’s AR industry specification group (ISG), says: “When the AR framework group was set up two years ago, the context was that, mainly, the solutions came from large American companies. There seemed a danger that some of the very good work done by smaller companies would not be able to find a way into the market.

“The ability for different providers components to communicate without need for conversion or translations we believe is good for scalability and a healthy technology ecosystem. It means that end users have confidence that they are able to evolve their systems when a better technology comes out. And they can swap these components for a new provider.”

With its heritage in mobile communications, ETSI is not the most obvious choice to try to set standards in AR, although the drive to make technologies like this work better over 5G means they are not as far apart as they first seem. One of the big selling points of 5G is in its much lower latency compared to the older cellular protocols. This should benefit AR and virtual reality because humans are very sensitive to motion lag in computer graphics that are meant to resemble and follow the real world, though there will always be issues with the finite speed of light when it comes to long-distance remote operation.

“ETSI is not necessarily one of the organisations recognised for expertise in AR. Its history has been in communications but is now addressing more verticals. The structure of the ISG allows us to get contribution from participants who are not members of ETSI,”  says Deschanel.

“At first there were five founding members, now up to 22. What was extremely good was that we have technology users as members. We have Bosch, Siemens and Schneider Electric among others, which has proved very useful for use-cases and requirements.”

The decision to start with a framework document is itself a recognition that the ETSI group is not trying to muscle its way into setting specifications in an area that already has a bunch of de facto choice and some other attempts at creating interoperable standards.

“We designed it specifically for augmented reality but the framework is applicable to virtual reality and any mixed-reality environment,” says Deschanel. “It gives a list of specifications that are relevant in general as some are not specific to AR. We aim to be comprehensive. We’ve launched a survey to collect more operational requirements: what types of conditions you use AR. A report on use-cases has been published, particularly industrial use-cases such as inspection, maintenance, training and manufacturing.

“What we wanted to achieve was to have an architecture that ecosystem partners and other standards organisations could refer to and look at. One of the tasks of the group is to map the specifications we have identified in the first report to the architecture, which is huge. There are many, many interfaces. There is no way the ISG could develop new specifications that cover it all,” Deschanel notes.

“We are working with a number of organisations to continue our work on that. We also plan to look at some interfaces that have existing standards that don’t completely meet the requirements of the architecture. And if there are no standards that match, then the group may start developing specifications if it has the expertise,” she says.

In terms of work with other groups, ETSI has signed an agreement with MPEG and a memorandum of understanding (MOU) with the Augmented Reality Enterprise Alliance (AREA). “We have a very close relationship with them and they have circulated the framework to their membership.”

Deschanel adds that one medium-term aim is to build an implementation of the architecture that will use open-source components either partially or in full to demonstrate the potential of a framework with interoperable components. She concedes there may be a problem getting buy-in from the bigger suppliers. “When you have a dominant position you are not necessarily in favour of standards,” she says, but adds that the group is focusing on users. If users start to demand AR to help run their operations but run into problems with making it fit their CAD and resource-planning software, they will be keen to tell their suppliers.