LiDAR systems are finding novel uses in today's world in anything from security to mapping to industrial automation. The mobility market is one area where there is an especially high level of attention and progress. LiDAR scanners are critical mechanisms in prototype systems for autonomous vehicles in addition to existing systems for adaptive cruise control (ACC), collision avoidance systems, traffic sign recognition, blind spot detection, and lane departure warning.
Not one of these LiDAR-based systems works without their sensors, their key component. This article focusses on how design engineers for original equipment manufacturers (OEMs) of LiDAR systems select amongst differing sensor technologies.
Distinguishing the environment quickly and reliably is an essential part of mobility light detection and ranging (LiDAR) systems in order to create as detailed a picture as possible of the immediate surroundings and the road ahead. Far ahead: systems mounted in fast-driving cars need to “see” at least of 150 meters forward, and perceive small objects down to 10 cm in height. The presents a technical challenge for the systems’ sensors.
Figure 1. LIDAR-scanner in action. Image Credit: First Sensor
The task requires harmonizing but independent sensor systems, with failsafe functional safety and environmental qualification. Units ought to be rated for operating temperatures from -40 to 125 °C (-40 to 257 °F) to entertain both environmental heating and heat from other system components, for example.
Sensors must have an optimum signal-to-noise ratio, to visualize the signal through any disrupting background. Optical detectors must be equipped to manage fluctuating levels of environmental light, therefore sensors should possess a wide dynamic range.
(Note the term “detector” may denote only the photoelectric detecting element, while “sensor” comprises the detector and adjacent electronics that deliver functions such as connectivity. The terms are occasionally used interchangeably.) Besides basic physics, LiDAR system designers should also contemplate basic economics. All components in the car should be greatly cost-efficient. For practicability, the best cost/performance ratio surpasses the best technology.
Currently, all automotive mobility systems utilizing long-range LiDAR are scanning devices which pass the laser beam gradually over the whole scene. Effective range using current technology is approximately 30 to 300 m.
Nearly all devices are constructed around 905 nanometer (nm) lasers: these emanate invisible beams, are accessible at low cost in high volumes, and utilize high power for short pulses (for example, 75 W peak for 5 nanoseconds), offering optimal power-to-cost ratio. These lasers have been widely used with mature, inexpensive silicon detector technology.
Selecting the Best Sensor Technology
As the industry progresses, design engineers are implementing many different sensor technologies for LiDAR mobility systems, each with its own has advantages and weaknesses.
Silicon PIN Diode Detectors
PIN diode detectors are silicon based and feature three semiconductor types layered together: P-type / Intrinsic / N-type. They display good dynamic range and can handle broadly fluctuating amounts of light. For instance, they can detect the reflection of a distant object, even when exposed to direct sunlight.
Although relatively inexpensive, they are unable to deliver the high levels of bandwidth or signal-to-noise performance most modern mobility LiDAR systems necessitate. Lastly, they are not very fast, or sensitive.
Silicon Photomultiplier (SiPM) and Single-Photon Avalanche diode (SPAD) Detectors
These solid-state, silicon-based sensors were initially developed for small, specialized scientific and medical applications but have more recently been trialed in the larger LiDAR market. Operationally similar to APDs (see below), these sensors are optimized for very high internal amplification or gain, meaning they can detect the slightest amount of light. Additionally, they are well-matched with commonly available CMOS technology, and can be paired with accompanying electronics on the same chip.
Whilst they are very fast, the sensitivity of their single-photon counters is much lower than that of APDs, so they depend on very high multiplication. Unfortunately, the multiplication method enhances noise that frequently damages the signal/noise ratio.
Their amplification mechanism is also susceptible to false triggers instigated by high temperatures. The most serious disadvantage of these sensors is that their high gain comes at the cost of saturation problems.
Initially, the sensors must manage laser light reflected from objects ahead. Several LiDAR systems require scanners with wide fields of view, which places quite a large amount of added light on an SIPM or SPAD sensor. LiDAR mobility systems regularly meet phenomena such as bright sunlight, high-beam headlights, or other LiDAR systems which can steep the sensor with higher light levels than it can handle, even when employing optical filters.
These sensors are often considered for various LiDAR applications because of the work to overcome these drawbacks. However, saturation issues and other difficulties cited above prevent them from being the detector of choice for scanning long-range LiDAR.
Figure 2. Comparison of detection technologies. Image Credit: First Sensor
Indium Gallium Arsenide (InGaAs) Photodiode Detectors
Often used at small sizes in telecommunications glass fibers, these sensors are newcomers to LiDAR, except for specialized military or aerospace applications. Conventional silicon-based construction is abandoned for InGaAs material.
This design ought to be more sensitive, and with a greater power output as the laser systems is specially built for its higher spectrum (1550 nm, versus 905 nm). The result is an automotive LiDAR system with a longer range than most other sensor.
However, InGaAs detector performance can be considerably degraded by even slightly higher than normal ambient temperatures: even in temperate climates, the sensor may require an external cooling system. Additionally, the base material is significantly more costly than silicon substrates and fabricating InGaAs sensors in large sizes necessary for LiDAR use would need much more complex manufacturing than silicon designs.
As yet, they have not been manufactured in high commercial volumes. Lastly, since this technology is new to the automotive LiDAR world, OEMs need to expend substantial time, effort, and revenue trying to develop a new LiDAR system around any InGaAs detector.
Avalanche Photodiode (APD) Detectors
Initially created for industrial and military applications, these silicon-based photodetectors enable incoming photons to trigger a charge avalanche, multiplying gain by their internal amplification mechanism. Their absorption-optimized structure translates at least 80% of a laser’s reflected 905 nm light into photoelectric current. The result is a greatly increased sensitivity.
Apart from their notable sensitivity, APDs possess an optimal signal-to-noise ratio, minimal saturation, and fantastic speed. They are amongst the lowest-cost sensor technologies obtainable.
One potential drawback is that APDs employ specific bipolar technology not well-suited with commonplace CMOS fabrication, meaning they can only be found in a small number of suppliers. They cannot be paired on the same chip with their allied CMOS electronics but experienced suppliers can engineer packages with sensor and electronics on closely adjacent chips, both of which can be augmented for best-in-class performance, without compromise.
An APD sensor array, for instance, can be accompanied by specially designed transimpedance amplifiers (TIAs) — with tailored gains and bandwidths — to translate the photocurrent to voltage, and to condition the signal entering the system for high gain. This capitalizes on performance, particularly in low-light situations.
APDs are created via an engrained, high-productivity commercial manufacturing method, verified in a wide range of systems already on the road. Essentially, when completed correctly, they combine proven performance with an attractive price. Presently the detectors of choice for automotive long-range LiDAR, APDs are essential in many of today’s most advanced mobility systems.
Figure 3. APDs. Image Credit: First Sensor
Selecting the Best Sensor Supplier
After determination of the ideal sensor technology, LiDAR system designers are still faced with choosing the correct sensor supplier. Candidates must be assessed cautiously: Do they have the technology, ability, and knowledge to alter their sensors and systems to an OEM’s individual requirements and markets? Will they work together with the OEM team on design, production, and timeline to ensure a winning time to market?
Insist on Experience
A LiDAR mobility OEM will lose the battle for quickest time to market if a sensor vendor has to bring its development, manufacturing, automotive qualification, and other processes up to speed.
Sensor suppliers attain experience by carrying out the work. A good contender will have applied their sensor/detector technology for mobility applications already. This may comprise standard and customized APD design; standard and customized dies, packages, and modules engineering and manufacture; and outstanding electronics.
Model suppliers will have a recognized performance record, with products such as automotive-grade APDs and connected electronics already utilized by major LiDAR OEMs.
Evaluate for Integrated Manufacturing
Priority for designers should be a supplier with pertinent technical advantages, including lowest noise and highest sensitivity, and a sensor maker that upholds holistic control of its field. The whole manufacturing process should be executed coherently, from processing the chips through to prefabricating sensor systems. By manufacturing all central components in-house, a supplier guarantees the long accessibility of all OEM products for series production and aftermarkets.
Check Customization Competences
Successful LiDAR system producers have the top cost/performance ratio; it can help separate a given system from the rest of the overloaded market. Off-the-shelf sensors may well not meet requirements and instead, components must often be tailored to exactly fit a preferred system design.
System makers must find an agile and responsive sensor supplier. Often, a supplier must collaborate with OEM designers to adapt the sensor and linked electronics for the tightest possible amalgamation with the rest of the system — and, consequently, for optimum performance.
Examples: the team must build sensor geometries suitable for a given selection of lenses; to optimize dimensions; and to otherwise adapt to the formations of each unique optical design. The team must deduce the optimum channel count — the number of signals received in parallel — to maximize the spatial resolution of the scanner. And it must tailor packaging for the shortest possible boundaries between sensor and electronics.
Lastly, a superior supplier should deliver sensors having refined technical advantages such as multi-pixel homogeneity. If photodiodes are not standardized and/or are diversely sourced, they will respond differently to ambient temperatures under real-world use. This can significantly damage LiDAR scanner performance. By contrast, multi-pixel homogeneity offers the tightest possible signal information distribution, even at peak distances.
Enquire About Automotive Qualification
Outstanding sensor suppliers should know the “rules of the road”: it should be knowledgeable in the newest automotive qualifications, toughness validation, and classification standards and regulations. Examples comprise ISO/TS 16949 automotive-certified production and testing, and AEC-Q 102 and 104 automotive-qualified APD array packaging.
The supplier should be capable of applying these and other applicable standards to all its components and manufacturing facilities, to observe regulations and to help system OEMs avoid liability. Growing regulation is unavoidable. Suppliers should show verified compliance through logged best practices, such as the rigorous self-qualification initiated by Ford Motor Company in its Q Program.
Search for Future-Proof Support
A supplier should exhibit a proven track record of quality and delivery performance, in addition to strong levels of support, from early development through to maintenance service. Sensor design should be factored in from the beginning of system design. The earlier an OEM includes the sensor supplier, the quicker and easier the entire design/manufacturing process turn out to be, and the better the ensuing LiDAR system performs.
Finally, a supplier should continually be looking towards future changes in this fast-paced field. The correct sensor creator will have an innovation roadmap of anticipated regulatory, business, and technological developments to come, to aid system developers in navigating this fast-changing market.
The sensor integral to every LiDAR system. System designers can select from many competing sensor technologies. Numerous design engineers feel that APD sensors deliver the best combination of performance and price. LiDAR system makers should also contemplate numerous aspects in choosing their sensor supplier — including experience, customization capabilities, and automotive qualification proficiency. As LiDAR and other mobility technologies continue to advance, making the right sensor choices visibly maps the path ahead.
This information has been sourced, reviewed and adapted from materials provided by First Sensor Inc.
For more information on this source, please visit First Sensor Inc.