Holograms herald era of virtual car design

17 Jun 2002

Ford could soon be going one step beyond virtual reality to test its new designs. In a joint venture with the UK's Defence Evaluation and Research Agency, it is replacing vehicle models with three-dimensional holographic displays. Michael Hatcher reports.

From Opto & Laser Europe July/August 2001

Western Europe's biggest science and technology organization, the UK's Defence Evaluation and Research Agency, is changing. This month sees it split into two very different operations. One half will become a public company known as Qinetiq, which will be devoted to commercializing innovations developed by the agency. The remaining half will continue to work for the UK's Ministry of Defence, as the Defence Science and Technology Laboratory.

Qinetiq will demand a more entrepreneurial, commercially minded regime than the old Defence Evaluation and Research Agency (DERA), and on the evidence of one optics-based venture, such an approach is already being taken. DERA is working with the US Ford Motor Company on a new way to view and design cars using a technology termed "replacement reality".

Car manufacturers currently use a combination of standard design-viewing software and clay models in the design process. Each time a design is updated a new clay model is made so that it can be assessed.

The idea of replacement reality is to build a three-dimensional system based around a holographic display that several designers or users can view and interact with simultaneously, without the need for the headgear or viewing goggles that are associated with virtual reality. Unlike virtual-reality images, a three-dimensional holographic image would be a real environment. Chris Slinger, team leader of the displays holography group at DERA's Malvern site, says that two key criteria must be met for this replacement reality to be effective.

First, users must be able to evaluate scale, form, texture and colour with confidence, which will demand a high-quality image and accurate depth cues - the cues that the human visual system uses to interpret images. Second, the model must allow natural and intuitive interaction with a number of users simultaneously.

Existing three-dimensional-display techniques, such as immersive and volumetric displays, go some way towards meeting these criteria, but, says Slinger, they cannot provide proper depth cues. He believes that holography is the only three-dimensional-display technology that supplies all of the depth cues that are used by our eyes.

For the purpose of car design, the hologram would be computer-generated, rather than captured on holographic film using the classic method. The computer-generated hologram (CGH) must then go through a light-modulation system to become a viewable image. Two bottlenecks have held up the development of three-dimensional CGH displays: the amount of computing power required, and the slow speed of the light-modulation technology available.

In a CGH, the interference between light waves is replaced by a mathematical calculation. The problem this poses to forming a computer-generated image is the amount of information that the CGH holds. For example, a 30 ¥ 30 cm plate classical display hologram contains more than 10 Tbyte of information.

However, such holograms have a resolution that is too fine for the human eye to resolve. Additionally, the horizontal spacing of a pair of human eyes allows a reduction in the degree of vertical parallax (the ability to look over and under the image) needed. This means that the number of pixels that the computer must display can be reduced to 10⁹ - which is a manageable feat.

The second problem concerns the electrically addressed spatial light modulators (EASLMs) that are used to update the three-dimensional image in real time to give the impression of reality. Although the modulators are fast, they have a low resolution and pixel count. Optically addressed spatial light modulators (OASLMs), however, tend to be slow, but have a high spatial resolution.

According to Slinger and colleagues, finding solutions to both of these bottlenecks is now realistic, thanks to new computer algorithms, hardware and novel modulators. Slinger and his team have developed the concept of "active tiling". In essence, this is a scalable system that exploits the high frame rate of EASLMs and the high resolution of OASLMs. The idea is that each EASLM output is optically replicated (actively tiled) on a section of a larger OASLM, which can then be updated at the video rates that are required to produce a convincing three-dimensional image.

Slinger admits that the decision to back this ambitious concept came as something of a surprise: "This is a bold move by Ford. We didn't really expect the idea to be taken any further than the preliminary research stage." He and his team are now working on a prototype system for the company, which is scheduled to be ready in autumn 2002.

In active tiling, the CGH is produced by a supercomputer and projected by an argon ion write laser (2 W, 514 nm) onto a two-dimensional array of EASLMs. Replication optics project a de-magnified EASLM image onto a bigger OASLM, which is capable of displaying at a high spatial resolution. By stacking the OASLMs next to each other, a sufficiently large display to produce images of the required size can be built up. Finally, readout optics form the holographic image.

The replication optics collect the EASLM image through a 5 ¥ 5 lens array and send it through a polarizing beamsplitter. Collimation lenses de-magnify the beam and project it onto the OASLM through another polarizer and a triplet lens array.

Slinger has been able to produce images of 2500 ¥ 2500 pixels at a pitch of 13 µm. However, the CGH pixels are not all coincident with the OASLM, so now Slinger is working on a 5000 ¥ 5000 field and a pitch of 6.5 µm to overcome this problem. So far, the active tiling system has relayed 6 Mpixel CGH patterns (from industry-standard CAD data files) written onto the OASLMs, and the images have been captured on a digital camera. A 670 nm diode laser replayed the images, but the low beam quality of this source degraded the image. Slinger is seeking alternative sources to achieve an improved readout quality.

As far as the readout optics are concerned, Slinger believes that a multimirror design is the most promising. This would reduce the maximum aperture needed in the optical system, although very large optics would probably still be needed.

Slinger believes that a practical electroholographic three-dimensional viewing system could be built within the next 3 to 5 years, assuming that there is a significant increase in computing power.

Support from Ford will be crucial - and it will be Ford's decision whether or not to build a full system in just over a year's time that will make or break the joint venture.

If the system does get the go-ahead, it could be just the first of many applications for such systems: other high-end uses could include aerospace design and data visualization for the biochemistry and chemical industries. It could also be one of the first successes of a new era for DERA. Visit DERA's Website for more details.