Aug 15, 2016 | By Alec

3D scanning takes many shapes and forms, but all take something from LiDAR (Light detection and ranging). This has been around since the invention of the laser itself, and uses light beams to find out exactly where pixels are in 3D space by measuring the returning signal. While widely used in robotics and even geographic mapping, it’s quite a cumbersome and expensive technology that doesn’t always produce results of an adequate quality. But the entire world of LiDAR and 3D scanning is about to be turned upside down, as MIT researchers are producing the LiDAR chips on 300-millimeter wafers, and which could can cost as little $10 to make. Most importantly, the non-mechanical beam steering in this device is 1,000 times faster than what is currently achieved in mechanical lidar systems.

This remarkable breakthrough technology has been revealed in an ieee article by PhD student Christopher V. Poulton, and his supervisor Professor Michael R. Watts, who are working with DARPA (The Defense Advanced Research Projects Agency) on photonic innovations. And as they argue, their new LiDAR chip will act as a disruptive force in the 3D scanning market and could be used in anything from robotics to autonomous vehicles, low-cost scanners and wearable sensors. Those fields are all regularly referred to as ‘the next big thing’, so this innovation could have a huge impact.

And as the MIT researchers reveal, their LiDAR innovation tackles the major limitations of existing systems. For most LiDAR systems (including those on autonomous cars) use discrete free-space optical components like lasers, lenses, and external receivers. In those setups, the laser is spun around while oscillating – limiting both the size and complexity of the scan and making it unsuitable for harsh environments. What’s more, they are expensive, costing anywhere from $1,000 to $70,000.

LiDAR on an autonomous car.

These new MIT LiDAR chips stand in stark contrast to those systems, as these chips could potentially be produced for $10 a piece (at production volumes of millions per year). “These on-chip devices promise to be orders of magnitude smaller, lighter, and cheaper than LiDAR systems available on the market today. They also have the potential to be much more robust because of the lack of moving parts,” they say. Most importantly, they are 1,000 times faster than current LiDAR systems and can allow an even faster scanning rate – very useful for tracking high speed objects from an autonomous vehicle, for instance.

Its origins can be traced to silicon photonics technology, which uses silicone waveguides a few hundred nanometers in cross section to create ‘wires for light’. Much smaller than optical fibers, they pave the way for photonic circuits on very small chips with properties similar to optical fibers except on a much smaller scale. These can be produced in CMOS foundries at large volumes, and address fundamental photonics issues such as waveguide loss and optical isolation.

This is something that has greatly interested DARPA, who created the Electronic-Photonic Heterogeneous Integration program (E-PHI) in 2011 with an eye on scalability. DARPA’s achievements, in which professor Watts was also involved, already demonstrated the commercial potential of CMOS foundries, and the MIT researchers saw the opportunity of applying the same manufacturing principles to LiDAR.

So how does it work? In a nutshell, these LiDAR chips are actually 0.5 mm x 6 mm silicon photonic chips featuring steerable transmitting and receiving arrays, as well as germanium photodetectors. While the laser hasn’t been integrated yet, that has already been demonstrated to be possible. To scan something, silicon notches inside a waveguide act as antennas that scatter the laser light into free space. Once out there, the LiDAR chips use the same time-of-flight measurement technique of existing LiDARs to scan something, but without being affected by sunlight or expensive photo-multiplier tubes and avalanche photodetectors. What’s more, no costly lenses are needed either.

Right now, this low cost LiDAR scanner can detect objects at a range of between 5 centimeters to 2 meters, though they are aiming for a 10-meter range within a year. “We have demonstrated centimeter longitudinal resolution and expect 3-cm lateral resolution at 2 meters. There is a clear development path towards LiDAR on a chip technology that can reach 100 meters, with the possibility of going even farther,” they argue.

The only problem right now is a limited steering range of 51 degrees, and overcoming that will be challenging because of size issues. However, they believe that a 100 degree angle should be attainable in the near future, meaning that a full 360 degree image can easily be obtained with multiple sensors.

So what’s next? Well, it’s clear that a lot of work still needs to be done and several material innovations could enable longer ranging and a higher lateral resolution. “The challenge here is how uniform and precise the silicon waveguides and antennas can be fabricated, and this capability will most likely increase in the future as lithography technologies improve,” they say. DARPA has already launched a program aimed at building on this silicon photonic LiDAR work.

If successful, these chips could break through numerous barriers for autonomous vehicles and robotics; multiple and inexpensive modules could cover vehicles and robots, giving them full vision of their surroundings. But you don’t need to be very imaginative to think of the numerous other 3D scanning opportunities that these chips could enable. While it can take a few more years before these LiDAR chips become commercially available, they will certainly change 3D scanning as we know it.

 

 

Posted in 3D Scanning

 

 

Maybe you also like:


   


mathew wrote at 11/17/2016 2:40:06 PM:

interesting point by Bill. What would look like the resolution in a 1 to 5 cm ranges???

Francis wrote at 8/16/2016 2:52:26 PM:

300 mm wafers are used for making a bunch of chips at once. Then they are sliced apart and make a lot of individual scanners. The issue for 3d printing is the resolution, which is very coarse. Light moves pretty fast, so it's tough to get high enough spatial resolution with low signal power and small optics. Really great technology, and this could have a very positive impact in many fields. But small 3D models are probably not going to be the hot application area.

Bill wrote at 8/15/2016 4:19:51 PM:

I was going to say the same thing... 300mm in size... that's 12 inches!!!

gregha wrote at 8/15/2016 3:41:40 PM:

300mm or 300nm???



Leave a comment:

Your Name:

 


Subscribe us to

3ders.org Feeds 3ders.org twitter 3ders.org facebook   

About 3Ders.org

3Ders.org provides the latest news about 3D printing technology and 3D printers. We are now seven years old and have around 1.5 million unique visitors per month.

News Archive