Dear Sadegh,

The following is only my personal point of view and does not necessarily reflects my company’s position.

As you, I sometimes have the impression that electric drives seem to be known and mature for quite a long time, with no major breakthrough. If you look at the industrial market, B2C or B2B with standard products, you cannot find solutions that seem so innovative. Would you like to build a robot or a drone, you can find a large amount of BLDC doing great work according to their datasheet. Problems arise when, for a new application, you cannot find a matching industrial solution. However, if you turn to the academic world, you realize the high number of publications and still some improvements and new topologies proposed. There is just a mismatch between what companies and researchers propose due to the new topologies not being industrialized yet, for not really clear reasons from the outside.

Take the example of robotics emerging markets. In some applications, the torque density needs to be as high as possible with the best back drivability possible (ideally direct drive). Addressing this application with traditional BLDC is complicated. You take a motor, you add a gear, but that’s not enough, that’s not what you want. Now, if you look at academic papers on motors made for torque, you will clearly find new topologies proposed: pseudo direct-drive permanent magnet machine in 2008, the resurgence of new type of Vernier machines from 2010, the same for doubly-salient PMSMs, introduction of partionned stators flux switching PMSM recently, etc. So, there is still research on topologies, there are still improvements. It’s not yet industrialized because remain the question of how to ensure constant performance and quality with low manufacturing cost when you need two or more moving parts, extremely small magnets to achieve high pole numbers, and so on.

This was only for the electromagnetic part. You have a lot of improvements made these days around bearings (and not roller bearings) with really amazing potentials. You can also find innovations in gears, in windings, better steel sheets and better magnets, high power and high speed electronics. The industrialization is still again the problem, these innovations are often touching the limits of traditional industrial machines. And even if doable for a prototype, it takes time to be able to do it for high volumes. New applications are really helping new developments since no conventional solutions are sometimes available. In some areas like [high torque with low speed] and [high speed with high power], you find a lot of developments utilizing new technologies.

I reply to your good remark. But it seems that it has not been saved. Have you received it?

Hey Sadegh, not at all received... I'm afraid you will have to type it again!

There is always a need for tailoring a mature technology for a specific case as you mentioned. This is usually done by the industry. On the other hand, a new type of application (say autonomous EVs) may need advancement of the technology. This can be achieved through research. I am emphasizing this type of need. It can be responded by using advancement in the contributing components of the technology; in this case new PM materials, new power electronic switches and converter topology, and new motor control methods.

A major step is to formulate requirements of emerging applications. For example, what the new robots are demanding from a motor drive system, which is not delivered at the present time. Industry can play a major role in this regard.

I agree with the major step you mention. And in this matter, companies are getting closer and closer to university laboratories, their spin-off, the startup ecosystem that's also often emerging from universities. That's the aim of my lab (Sonceboz) at the EPFL (Lausanne-Switzerland) for example, and Maxon Motors is doing it too. It's hand-in-hand collaboration, sometimes open-innovation. For the records, the two companies I mention are working on exoskeleton motors, both in their own ways (Twiice for Sonceboz http://twiice.ch and other drive for Maxon https://www.maxonmotorusa.com/maxon/view/news/Exoskeleton-joint-actuator), but both help define the requirements with laboratories/startups/companies who are not specialists of motors.

This type of collaboration would be very fruitful for all parties if it is managed well in a prepared industrial and social atmosphere. However, it is sometimes difficult to fulfill the desires of all players. There is no standard for such a collaboration, at least as I am aware. Would it possible to be more specific about your lab at EPFL if you like. What would be the win-win play for the company, the university, the profs and students? What is the input and outcome for each or for all?

Yes, there are no standards. You can find many recognized papers and books about innovation with the outside of the company but there is not a standard way of doing, as it's too much dependent on many factors.

The laboratory develops a new technology, they have the knowledge around it, might have already created design tools and prototypes. A company interested in it would like to obtain this technology. They can either pay for the patent that the lab has created, or they can work with the lab to make technology transfer. In case the technology is not mature, you cannot directly make a technology transfer and the laboratory needs time which means PhD students, which means money. In all the cases, the company brings money in different ways to the laboratories (paying for a 4-year PhD, a 6-months internship, a 1-year postdoc project). To fit better the needs of the company and to make interesting/useful research, the two parties (lab and company) work jointly to make the technical research an innovation (to be marketed).

• company: learn in detail about the technology, land innovation, make contacts with lab and other arounds (to scout for other tech), …
• university: their tech is used in commercial products, lab is partly paid from external sources
• prof: the tech can lead to a spin-off, get resources from the company to develop a realistic innovation, make contacts with companies (to scout for other ideas), …
• students: get financed, applied work, contacts for job scouting and learning

You provided a brief and clear picture of your model. Hope it works well, with good results for all parties and advances the technology and innovation.

Our team works on improving performance of PMS machines under advanced and practical control systems; and also on wireless power transfer by a total of 8 full time graduate students and a number of part time graduates.