IoT news of the week for Oct. 9, 2020

Here’s a privacy-centric way to track employees in buildings: Density is launching a product called Open Area that uses radar sensing to track people moving inside buildings. Density may be familiar to y’all as a company I’ve covered due to its ability to count the number of people entering and exiting a building. It does that using lasers and time-of-flight sensing from a device attached to a doorway. Now it is adding the ability to count people and see where they are in a larger area independent of doorways. The Open Area product is exciting because it’s a base component necessary for understanding where people are in a space and how they are using that space. The technology tracks a person by monitoring clusters of points that appear in a room, but it can’t tell if someone is a man or a woman, their age, or anything that could be used later to identify them. Density CEO Andrew Farah calls this privacy by design, and says it’s essential to getting companies to use the Density product. “Most employers don’t want to use cameras to track their workers,” he says. With Open Area, a company can see where people are, how they use a space, and then make decisions about office space and office configuration. (Density)

Emerson will acquire Progea to visualize factory data: Emerson said this week that it will acquire Progea Group for an undisclosed sum. Progea will become part of Emerson Automation Solutions, and its industrial IoT software will help Emerson add factory visualization and provide flexibility in managing programmable logic controllers (PLCs). Essentially computers that control a machine, PLCs are the heart of industrial processes. They are typically on their own closed networks; any data that comes off them is then monitored or fed to IT networks for maintenance or other IoT use cases. (TedMag)

Amazon Halo’s principal medical officer has a dystopian mission: This article interviews Dr. Maulik Majmudar, principal medical officer for Amazon’s Halo wearable device. It focuses on his assurances that the Halo device will respect user privacy, that its ability to detect a person’s “tone” using their voice is going to be helpful. But I was struck by his characterization of his job, which is to “add clinical credibility to the features within Halo.” The phrase “add clinical credibility” — as opposed to, say, “clinically validate” — makes it sound like this man’s job is to put a gloss of medical endorsement on the product. And indeed, he makes clear in the interview that this product is not a medical device. He suggests instead that folks think of it more as a device designed to encourage healthier behaviors, noting that a doctor was involved in its design. There’s a lot of digital snake oil in the wearables and consumer wellness market already, so I suppose it’s nice that Amazon isn’t trying to mislead us. (Protocol)

What the heck is UWB and why should IoT care? Kevin Tofel tackles the potential of Apple’s ultra-wideband chip and Google’s Soli for delivering context for the smart home and helping us control it. (StaceyonIoT)

Google offers a good example of ML at the edge: One of the most common questions I get about machine learning at the edge is, what? As in, what can we do with it? Google is providing a really cool example with the launch of a feature for the deaf and hard of hearing on Android devices. Android phones will have the ability to listen and identity 10 different sounds locally on the device. If one of these sounds is detected, the phone can offer a notification or vibrate to let the person know that a baby is crying, water is running, or glass is breaking. This is a cool accessibility feature, but it’s also an answer to the question of what you can do with local machine learning at the edge. (CNET)

Electric truck company gets $10M: Einride, one of my favorite startups in the autonomous vehicle sector, has raised $10 million for its electric trucks. Einride makes electric, autonomous trucks designed to operate under the oversight of a human. The idea is that one person could monitor a few trucks and offer oversight and help as needed, much like pilots currently fly drones in the world’s militaries. I like that Einride has been testing its trucks successfully in Sweden for the last year or two, and that is trying to take an innovative approach to a small part of the distribution and delivery chain. (Crunchbase)

People aren’t robots and we shouldn’t expect them to be: This story breaks my heart because it embodies everything awful about how certain companies are approaching IoT and AI to cut costs and boost efficiency. The Center for Investigative Reporting obtained memos and other documents that show how Amazon’s workers reported more injuries as the company installed more robots in its warehouses. It wasn’t the robots that made the warehouse more dangerous, but the push by managers to force humans to keep up with the robots. It is true that the addition of robots to a fulfillment or production line speeds everything up until a human gets involved. But a responsible way to manage this challenge isn’t to ask humans to boost their quotas to meet the robot’s production. You either need to design humans out of the line or redesign the line so the number of humans at the bottleneck rises, allowing them to handle the increased efficiency of the robots. Automation isn’t simply about replacing humans, but augmenting them, which means that adding automation will change the way we use humans and deploy them. (Reveal)