From texting on a smart phone to ordering train tickets at a kiosk, touch screens are ubiquitous and, in most cases, relatively reliable. But for people who are blind or visually impaired and use electronic braille devices, the technology can be vulnerable to the elements, easily broken or clogged by dirt, and difficult to repair.

By combining the design principles and materials of soft robotics with microscale combustions, Cornell researchers have now created a high-resolution electronic tactile display that is more robust than other haptic braille systems and can operate in messy, unpredictable environments.

The technology also has potential applications in teleoperation, automation and could bring more tactile experiences to virtual reality.

The research is published in Science Robotics. The paper’s co-first authors are Ronald Heisser, Ph.D. ’23 and postdoctoral researcher Khoi Ly.

“The central premise of this work is two-fold: using energy stored in fluid to reduce the complexity of mass transport, and then thermal control of pressure to remove the requirements of complex valving,” said Rob Shepherd, the John F. Carr Professor of Mechanical Engineering in Cornell Engineering and the paper’s senior author.

“Very small amounts of combustible fuel allow us to create high-pressure actuation for tactile feedback wherever we like using small fluid channels, and cooling the gas during the reaction means this pressure stays localized and does not create pressure where we do not want it,” he said. “This chemical and thermal approach to tactile feedback solves the long-standing “Holy Braille’ challenge.”

The majority of refreshable electronic tactile displays contain dozens of tiny, intricate components in a single braille cell, which has six raised dots. Considering that a page of braille can hold upwards of 6,000 dots, that adds up to a lot of moving parts, all at risk of being jostled or damaged. Also, most refreshable displays only have a single line of braille, with a maximum of roughly 40 characters, which can be extremely limiting for readers, according to Heisser.

“Now people want to have multi-line displays so you can show pictures, or if you want to edit a spreadsheet or write computer code and read it back in braille,” he said.

Rather than relying on electromechanical systems—such as motors, hydraulics or tethered pumps—to power their tactile displays, Shepherd’s Organic Robotics Lab has taken a more explosive approach: micro combustion. In 2021, they unveiled a system in which liquid metal electrodes caused a spark to ignite a microscale volume of premixed methane and oxygen. The rapid combustion forced a haptic array of densely packed, 3-millimeter-wide actuators to cause molded silicone membrane dots—their form determined by a magnetic latching system—to pop up.

For the new iteration, the researchers created a 10-by-10-dot array of 2-millimeter-wide soft actuators, which are eversible—i.e., able to be turned inside out. When triggered by a mini combustion of oxygen and butane, the dots pop up in 0.24 milliseconds and remain fixed in place by virtue of their domed shape until a vacuum sucks them down. The untethered system maintains the elegance of soft robotics, Heisser said, resulting in something that is less bulky, less expensive and more resilient—”far beyond what typical braille displays are like.”

“We opted to have this rubber format where we’re molding separate components together, but because we’re kind of molding it all in one go and adhering everything, you have sheets of rubber,” said Heisser, currently a postdoctoral researcher at the Massachusetts Institute of Technology. “So now, instead of having 1,000 moving parts, we just have a few parts, and these parts aren’t sliding against each other. They’re integrated in this way that makes it simpler from a manufacturing and use standpoint.”

The silicone sheets would be replaceable, extending the lifespan of the device, and could be scaled up to include a larger number of braille characters while still being relatively portable. The hermetically sealed design also keeps out dirt and troublesome liquids.

“From a maintenance standpoint, if you want to give someone the ability to read braille in a public setting, like a museum or restaurant or sports game, we think this sort of display would be much more appropriate, more reliable,” Heisser said. “So someone spills beer on the braille display, is it going to survive? We think, in our case, yes, you can just wipe it down.”

This type of technology has numerous medical and industrial applications in which the sense of touch is important, from mimicking muscle to providing high-resolution haptic feedback during surgery or from automated machines, in addition to increasing accessibility and literacy for people who are blind or visually impaired.

“As technologies become more and more digitized, as we rely more and more on computer access, human-computer interaction becomes essential,” Heisser said. “Reading braille is equivalent to literacy. The workaround has been screen-reading technologies that allow you to interact with the computer, but don’t encourage your cognitive fluency.”

Amazon’s budget security brand, Blink, announced two new cameras during the company’s fall hardware event in New York City: the Blink Mini 2K+ and the Blink Outdoor 2K+. As the names suggest, these cameras sport 2K resolution to pick up more details. But it’s the Blink Arc accessory that makes these security cameras more interesting.

The Arc can combine two cameras and stitch the feeds together for a complete 180-degree view. The product of some glue-gun experimentation, a borrowed 3D printer, and some nifty AI tools, this accessory came together in just 60 days. “There’s this fatal flaw with pan/tilt cameras,” says Jonathan Cohn, head of product at Blink. “The cliché Mission Impossible scene, where they wait for the motorized pan/tilt to turn the other way and duck behind.”

Camera Fusion

Blink’s Arc is Cohn’s solution, initially cooked up in his kitchen, and it’s designed to eliminate your blind spots. He showed some photos of the early prototype, cobbled together using snap mounts and hot glue. He was able to get the right angle so that he’d walk out of the frame on one camera mounted on the front of his house and into the frame on another camera on an almost level horizon.

The junior mechanical engineer, tasked with perfecting the angle and necessary overlap, used an AI tool to stitch the videos together, and despite some warping, it looked promising straightaway. Cohn borrowed his kid’s 3D printer to build the first Arc and popped in a couple of third-generation Blink Mini 2K+ cameras. Before long, the computer vision team found a way to dewarp the video, and the Arc was producing an almost seamless 180-degree live view.

Part of the charm of Amazon’s budget security camera brand is the jerry-rigged nature of add-ons like the Blink Mini Pan-Tilt, which allows you to slot a Blink Mini camera into a base that adds pan and tilt functionality. The Blink Arc is very much in the same mold. You can slot in your existing Blink Mini 2 cameras or snag a couple of the new Mini 2K+ models. (It doesn’t work with Blink’s other devices.) As long as the cameras are the same type, the Blink app can stitch their views.

The Arc houses the cameras at just the right angle and enables them to use a single outdoor power supply. You can connect the Arc to a snap mount to set it up horizontally, vertically, or hanging under your eaves. The video stitching is done on the software side; you simply link the cameras in the Blink app as left and right, and get a panoramic view. Blink even worked out a way to pan and zoom on subjects, so when there’s an event, it looks much like a pan/tilt camera tracking a subject, but it’s really just zooming on the 180 feed.

App-based companies have publicly spoken for years about the money-saving potential of autonomous vehicles. These firms have poured billions of dollars into recruiting and managing the independent contractors who do the delivering and driving for them, and millions more ensuring they’ll stay independent contractors and not employees. What if the firms could skip all that? What if robots did all the work, or at least some of it?

Still, with today’s announcement, DoorDash is throwing its lot into an industry that has faced some choppy waters. And, of course, the threat of public kicks.

Speed Bumps

Delivery robots were hyped during the onset of the Covid pandemic as a solution to that other very human problem of contagion. Since then, however, Amazon and FedEx abandoned their delivery robot projects; others working on delivery bots have pivoted to software or industrial uses. The companies that remain have mostly focused on smaller deployments on college campuses or a select few cities, and those don’t seem to be growing as quickly as hoped.

Estonian company Starship Technologies, the biggest one still standing in the delivery robot space, has found a niche operating on mostly university campuses, where streets and sidewalks are wide, well-maintained, and relatively friendly, and where those desperately seeking 2 am pizzas and burritos are at their least price-sensitive. Postmates spinoff Serve Robotics launched in 2017 but has built only 400 robots, according to its most recent financial flings, with goals to build 2,000 by the end of the year.

Contrast that with the growth in autonomous vehicles. Though robotaxi services are still limited to a handful of global cities, they’re picking up and dropping off customers to the tune of hundreds of thousands of rides per week.

The reason for the slower growth in delivery robots is actually pretty simple, says Bern Grush, the executive director of the nonprofit Urban Robotics Foundation: “You’re trying to solve a much harder problem with far, far, far less capital and far, far, far less compute.”

Consider the technical challenge Dot has ahead of it: DoorDash says the robot is built to operate on sidewalks, bike lanes, and roads. It’s meant to pilot in and out of parking lots to pick up food and to navigate driveways and apartment complexes to drop it off. That means the software needs to “understand,” predict the movement of, and get around a remarkable number of situations, vehicles, and living things: cars, trucks, school buses, strollers, children’s bicycles, aggressive mopeds, motorized wheelchairs, dogs, squirrels, toddlers, people with limps, runners, competitive cyclists. And on and on.