Don’t curse the elements the next time you’re confronted with a day of freezing rain and you need to run an errand. Think of it as your chance to demonstrate why mankind still has the edge over machines through the wonders of sensor fusion. For now, anyway.
It’s rainy, it’s freezing, and walking is risky business. Our combined senses provide us with the ongoing ability to navigate, or step, cautiously. If we had to individually process and analyze separate sensory data inputs, the delay might cause us to injure ourselves by slipping on a frozen sidewalk before we realize that not only is it raining heavily, but the rain is freezing and the sidewalk is slick. In concert, our abilities of sight, hearing, and touch let us instantly, unconsciously, take stock and act appropriately – our natural sensor fusion. So pat yourself on the back – after you’re indoors, sipping your favorite hot beverage, and catching up on the day’s headlines – for the natural sensor fusion powerhouse you are.
As for that article you just spotted about driverless cars in your favorite daily paper, weekly news magazine, or online – here’s why sensor fusion matters there as well. Essentially, cars will need to collect and process huge amounts of data as fast or faster than humans and they’ll do so via sensor fusion.
Getting this technology working effectively and determining the optimal suite of sensors and software will be critical to ensuring that vehicle automation can actually bring major benefits to American transportation. Vehicle automation potentially will only increase safety, improving mobility for travelers, and increase energy efficiency while simultaneously reducing pollution. Last, but not least: driver error is a contributing factor in more than 90 percent of all crashes and costs the U.S. economy more than $200 billion per year in medical, property, and productivity losses, according to the Subcommittee on Highways and Transit of the U.S. House of Representatives.
But how does sensor fusion work?
Radar, odometers, GPS – these devices may be familiar to you. Each is an example of a wide range of sensors, including inertial navigation systems like those developed and manufactured by KVH, that can be incorporated into unmanned and autonomous vehicle platforms, yielding data needed for accurate and trustworthy navigation, guidance, and control.
Not only that, the computational platform needs to not only mimic human sensing but perform better than a human not only in ideal road conditions but in compromised environmental conditions – fog, freezing rain, etc. – or situational anomalies like potholes or another vehicle backing out into the road that could fool the platform and put the vehicle in jeopardy.
So far, so good: Scientists and engineers are close to making that happen. Advances in sensor technology and processing techniques make it possible to combine and integrate fiber optic gyroscopes, accelerometers, magnetometers, pressure sensors, optical systems, and other sensing devices with onboard processing and wireless connectivity to begin to emulate the natural data fusion that occurs in biological systems.
In other words, sensor fusion can deliver a new, more integrated and efficient – almost human – way to sense position, direction, and attitude in navigation and control systems, and it’s increasingly considered key to the successful development of fully autonomous systems.
It’s all worth pondering, especially when the next errand confronts you on a cold, rainy day – and you’d rather not go out.
Discover the challenges and solutions to making unmanned and autonomous platforms as safe and effective as possible. Download the free ebook, “Unmanned Technology Challenges: Teaching Machines to Operate Like Humans” today.
For a detailed explanation of the technology involved, take part in our free webinar, “Sensor Fusion: Potential, Challenges and Applications,” available on demand for your convenience.