Microsoft had hit hard in January 2015 by unveiling HoloLens, an augmented reality headset whose performance was remarkable at the time. Its successor, the long-awaited HoloLens 2, was announced in February 2019 and was able to win against the Magic Leap One. Despite these initial successes, the future of the product now seems very uncertain within Microsoft.
A Business Insider investigation published on February 2nd reports internal dissensions and even the cancellation of the project (called Calypso) that was supposed to lead to HoloLens 3. It is based on the testimonies of three collaborators involved in the project. The Digital Factory can corroborate this information that it also obtained from an internal source several months ago.
A denial that is not one
Microsoft was quick to respond through its VP of Communications, Frank Shaw, and Alex Kipman, who leads the teams working on AR/VR. They insist that Microsoft is continuing its “metaverse” efforts and that HoloLens is one of them. Obviously, they do not directly contradict the fact that a HoloLens 3 development project was canceled in 2021. A classic technique consisting in “denying without denying” to reassure customers and partners.
Thus, it is quite possible that the company will release a product named HoloLens 3 in the future, but it will be different from the Calypso project that Microsoft put an end to in mid-2021. The latter ran under a modified version of Windows, like the first two HoloLens, and was designed to be more resistant, to be able to be used outdoors (HoloLens 2 does not work in too bright an environment) and to have a better battery life.
It is likely that the development of the Calypso project was related to the needs of the IVAS program of the United States Army, for which Microsoft won a juicy contract of 22 billion dollars (mainly for cloud services) but which has been undergoing significant delays since.
Three weeks ago, the Wall Street Journal pointed the finger at the fact that about a hundred engineers left Microsoft’s “mixed reality” division in 2021 (it has about 1,500 people in total). Almost half of them left for Meta, while others joined Google or Apple. Many were veterans of the company, and all of them were working on the upcoming HoloLens.
As for the reasons for its discontinuation, they may have their sources in the problems inherent in the HoloLens 2. We were at its launch at the Mobile World Congress 2019, and the promises made on stage were exciting, especially regarding the greatly extended field of view compared to the first model. However, it soon turned out that the display quality had suffered greatly from this increase in area, which was also not as large as advertised.
At issue: the Laser Beam Scanning (LBS) display technology chosen to obtain this wider field of view. In addition to a lower resolution than on the first HoloLens, our many uses of the headset have frequently been accompanied by annoying chromatic aberrations, the image becoming for example almost monochromatic in some cases. If Microsoft has been stubborn in this direction, it is possible that a failure to advance this technology is at the heart of the abandonment of the project.
A partnership with Samsung
If Microsoft has put an end to Calypso, the company is not abandoning the market for all that. Also according to Business Insider, the resources would have been redirected to the Bondi project, launched in mid-2021, which is based on a partnership with Samsung. It would be a very different device whose principle is similar to that of Meta’s Cambria project and Apple’s future headset.
Rather than projecting images onto transparent glass (by waveguide), it would use cameras on the front of the helmet to capture the environment and then retransmit it on screens. So it would be a virtual reality headset capable of doing augmented reality. The term used in the industry for these dual-use devices is “mixed reality”, but (and this is often confusing) Microsoft has been using it from the start to talk about HoloLens. An abusive use, just like that of the term “hologram” for that matter.
Samsung and Microsoft have already collaborated in VR with Odyssey headsets. Note that the computing power would not be integrated into the headset but provided by a Samsung smartphone. Another project, still in the embryonic stage, would consist of a headset that works by connecting to the cloud. In summary, with the abandonment of the Calypso project, Microsoft seems to have drawn a line under the idea of an all-in-one device running Windows. This means, among other things, that the applications developed for HoloLens 2 may not be compatible with the next devices.
Uncertainties about the strategy
As a result of these turnarounds, the morale of the teams would be at its lowest, according to about twenty employees and ex-employees that Business Insider interviewed. They describe a general confusion and uncertainty about the strategy to be followed, especially as factions clash over the course of action to be taken. Some would like to persist in developing their own hardware internally, while others think it is better to focus on the software to create the reference tools of the metaverse (like Mesh, which is far from mature). Satya Nadella, the CEO of Microsoft, would lean towards this second option.
The abandonment of HoloLens in favor of a partnership with Samsung is in any case not only happy, hence the resignations. The target market is also a source of conflict between those who think it is necessary to focus on professional (and military) uses and those who would rather attack the general public. This would be the case in particular of Rubén Caballero, an Apple alumnus recruited in 2020 to manage hardware engineering under Alex Kipman. In summary, despite the recurring mentions of the importance of the metaverse in the company’s plans, Microsoft would not have a clear roadmap in this matter. Alex Kipman is challenged on this point by the testimonies collected by BI.
A great vision… and poor execution
This not very reassuring observation gives an impression of a mess because HoloLens was really impressive when it was released and has been able to convince prestigious customers over the years. Unfortunately, this scheme is also a great Microsoft classic: to position itself very upstream on a disruptive technology with an ambitious vision and often an imperfect but promising prototype or first product… then fail to perfect the value proposition and end up getting double-crossed.
We have seen it in “tablet PCs”, in proto-smartphones running Windows Mobile, with Zune and Kin, with Kinect… There is no shortage of examples, up to the fiasco of the late Windows Phone or even the recent Surface Duo, without forgetting of course Windows on ARM. And what about the Windows Mixed Reality environment, integrated by default into Windows 10 but with which nothing has been done? Microsoft again seems unable to transform the trial.
It’s anecdotal, but we had this same conversation with a member of the HoloLens team at Microsoft after the release of the second model. The person has since left the team in question.
One of the great difficulties of virtual and augmented reality technologies is that technological constraints are still very strong today. The entire sector has been driven forward for the past five years by Meta, which is putting its whole future into it and is spending incessantly to ensure that it is a leader there. It must be said that he does not really have a choice: to free himself from the yoke of Apple and Google, who rule over smartphones, it is necessary to cut the grass under their feet on the next paradigm shift.
Microsoft is in a relatively similar situation: a former king of personal computing from the time of PCs, it got overwhelmed in smartphones and its Windows business is no longer the engine of its growth today. However, he is not ready to invest heavily like Meta, the temptation being to focus on activities that are already generating money, such as the cloud, and whose growth and sustainability are assured.
A reasonable but double-edged logic, because half an effort does not usually produce great results in this environment. Google suffers a bit from the same problem, with a mixed reality headset project that it is impossible to say today if it will actually see the light of day, and which risks in any case arriving late in front of Meta and Apple. So it is difficult to have confidence in Microsoft’s plans in this matter until further notice. By the way, this debacle represents an opportunity for Magic Leap, which could impose itself under the new leadership of Peggy Johnson, a former executive at Microsoft.