[interview] The secret to boosting innovation at b<>com: Advanced engineering

Philippe Lemonnier, Director of Engineering & IT

Fred Pieau
Going from an innovative concept birthed in a research lab to a technological solution that can be transferred to a client in record time? That’s b<>com’s mission. It’s why the IRT has a Department devoted to software and hardware engineering as well as IT.

Although the vast majority of research organizations don't have their own in-house engineering division and need to seek outside contractors to begin the process of bringing an idea to market, b<>com is quite an exception. We’ve decided to spotlight this work by interviewing the director of that division, Philippe Lemonnier, who has a 30 years’ experience in the field.

Philippe, how is this engineering team organized?

Our department has about thirty engineers, most of them recruited from industry. This was something we looked for at the time of their hiring; we were trying to find people experienced in bringing products to market so that the research projects could benefit from their assistance. In terms of how the team is broken down, on one side are the IT and software engineering teams, both led by Pascal Gravoille, and on the other, we have a hardware engineering team headed by Benoit Le Ludec.

What's the strength of this engineering team?

Definitely the variety of highly developed skills, which when combined, are suitable for providing guidance on a fairly wide range of research projects from cybersecurity to artificial intelligence, from 5G to virtual and augmented reality, and also including cognitive technologies and image/audio processing. This diversity of skills has also enabled us to build our infrastructure atop our very own datacenter, created from scratch by the crew at b<>com. And that's a pretty big deal! You need an incredible variety of talents to get a thing like that off the ground. Advanced knowledge of thermodynamics, energy, and acoustics are just some of what’s needed... not to mention the obvious computing and networking skills. A whole suite of capabilities that most people wouldn’t even realize was necessary. Overall, it’s important to understand that the Software teams work in tandem with the Hardware and IT teams; this forms a continuum of skills that are essential for turning research projects into a commercial reality.

At what point do the engineering teams get involved in the research projects?

Ideally, as early as possible. For instance, some team members are on the project incubation committees, if only to be aware of what’s on the drawing board. They can also guide or advise the project leaders on certain aspects related to industrial standards or the best candidate technologies for implementation, which is not something that exists in the world of academic research, to take one example.

In some respects, this might look similar to what the industry usually does with R&D teams backed up by an engineering department, but in our case, we have reimagined the process to move even faster by not linking serially early-stage research and engineering but rather work in parallel, in order to meet the market’s technology needs within a much shorter timeframe.

In concrete terms, how do you move from concept to tangible to transferable at b<>com?

This requires the involvement of a multitude of skilled players: First, research labs, because for each project, researchers bring their concept to a certain level of maturity so that a demonstration can be provided. The engineering approach is mostly meant to raise the TRL (Technology Readiness Level) of that proof of concept. This is done methodically, with an industrial approach, in order to obtain a transferable solution with a real value, which a third party can master and quantify. Engineering must essentially answer these questions: What do we need to do so that this innovation can be put into the hands of our clients and prospects? What are the most effective, suitable software and/or hardware solutions that are truly transferable because they are simple to integrate at the appropriate cost of implementation for the product they will be in? Will they function properly over time? Do they comply with legal and regulatory standards?

b<>com has 3D printers, so it can make all sorts of parts that don’t necessarily exist on the market.

Fred Pieau

Can you give us an example success story to illustrate this work?

One of b<>com’s recent success stories is *Adaptive HDR Converter* which is an innovation that has won several awards at international conferences and is now used by many major players in the world of broadcasting. This concept came straight out of our labs, with a clever algorithm developed to meet the needs of our clients, while also being as simple as possible to integrate into their own final product.

And for a smart solution, you need to come up with an architecture that has the proper balance between hardware and software in order to achieve something compact, inexpensive, and therefore easy to integrate into the client’s product. As such, the engineering phase can also be a source of innovation. In fact, the inventive solutions, sometimes very original, used to give life to our researchers’ innovations are themselves yielding a separate category of intellectual property.

After designing the electronic circuit diagrams, we move to the physical stage! From small, fairly simple components to more complex assemblies, the Hardware team has a wide range of talents.

Fred Pieau

One of the team’s latest "creations" is the miniaturized version of the *Wireless Edge Factory* which ended up as the smallest 5G box in the world. How and why was it designed?

This box was created in order to provide demos at conferences & tradeshows to demonstrate how easily our software technology *Wireless Edge Factory* can be embedded and deployed in any hardware. It was designed with the support of a variety of skills, including 3D printing and hardware integration. The hardware aspects involve knowledge of electronics, mechanics, thermodynamics, acoustics, and electrical safety. Designed to catch the eye, its elegant style is ultimately only the tip of the iceberg, which is meant to highlight the work of the Advanced Connectivity laboratory and reflect the fact that b<>com is able to provide software technologies that our clients’ 5G applications can run on. The software and hardware packaging work was therefore conceived at an early stage and shown off at conferences, but naturally, the main step that comes next is PoC (Proof of Concept) so that our clients can actually try out the solution before they adopt it.

At left, the 2019 version of the *Wireless Edge Factory*: 50kg in 85*53*30cm, exhibited at IBC 2019. At right, today, the new miniaturized version that is 25*24*30cm and weighs 6kg can fit into a carry-on bag on a plane!

Fred Pieau

Confidentiality prevents us from disclosing our current projects, but can you tell us in a few words what might be coming soon?

One of the most promising and innovative subjects that comes to mind is a very early-stage technology from the New Media Content Lab. Named b<>com [Smart FPS], it’s an algorithm that dynamically determines the minimum frame rate that can be achieved in a video without losing quality. It represents bandwidth savings of up to 40% during the production phase and 15% during transmission. This technology came out of the lab and was presented at several shows, including IBC, in 2019. It made a strong impression in the broadcasting industry! Very soon, this algorithm is expected to take the form of either software or hardware or possibly a combination of both, which a client can easily integrate. The many types of expertise that must be assembled to achieve this make it a passion project for the engineering team. It’s an even bigger challenge than the SDR-HDR solution, but more than that, it's an opportunity to bring our clients a new signature b<>com technology!

>>> interested in a demo of our solutions? Please contact sales@b-com.com