Robotics

With drug development front page news around the world, a quieter story is spreading in automation circles: The rise of robots as indispensable lab tools in drug development. As the need for faster drug development surges, manufacturers and researchers have adopted systems that promise new levels of efficiency and accuracy.
My colleague reported on the latest example this week, an AI platform and automated drug lab from IBM, which could cut the cost and time associated with drug discovery and synthesis. The IBM system uses computational modeling and predicts chemical reactions in order to aid human chemists in the creation of new molecules. The process allows biochemists to work remotely via an automated lab.

A chemist could be sitting at home who was willing to make a molecule, and after connecting to RoboRXN for Chemistry via a web browser, they draw the molecule.
RoboRXN would then recommend optimal scientific routes and the best starting material available commercially. 
Once submitted, RoboRXN would self-program itself to “execute the process in an autonomous laboratory.” In other words, experiments could be conducted remotely with the right integration and hardware setup.

It’s part of a broader trend toward automation in the hard sciences, which, somewhat ironically given their scientific bent, have been slower than many other industries to embrace automation technology. The barrier to automation has long been a difficult combination of intense regulation, budget constraints, and a trade tradition of seasoned craftsmanship on the lab bench, all of which have slowed the adoption of lab automation.
There has also long been a labor case against automating laboratories. Grad students and interns work impossible hours for little or no pay in academic and clinical settings, where future scientists and researchers learn the ropes and much of the science underlying drug development originates. And because labs often fall under the managing jurisdiction of a highly-credentialed and invariably busy person, it can be difficult for companies to market their products to anyone with decision-making capabilities.
The coronavirus has upended that paradigm. With barriers to working in person, many research institutions are left scrambling to find a way to do hard science. Essential worker exemptions have allowed many drug manufacturers to continue working (a Gilead facility in the south bay region of Los Angeles is thrumming with activity these days; Gilead manufactures Remdesivir, which is being used in some COVID-19 treatment protocols), but the lockdowns have only accelerated the glance toward automation in the hard sciences, a trend seen in industries as disparate as fast food and grocery delivery.
That realization coincides with a decade of falling sensor prices and increasing market penetration of robots, which have brought the costs of automation technology down. That’s led to spiking interest in automating tasks like liquid-handling, which is a notoriously error-prone component of work in the hard sciences.
A Swiss company called Andrew Alliance, for example, makes liquid-handling robots for life sciences labs and has raised millions in backing. Andrew Alliance’s flagship product is a tabletop robot that fits in a footprint about the size of piece of notebook paper. Thanks to advances in collaborative automation, the robot is ready to go out of the box and is easy to operate.
“Demand for better reproducibility continues to increase,” said Otello Stampacchia, managing director and founder of Omega Funds, an early investor in Andrew Alliance.

Another company called Opentrons makes a pipetting robot for use in labs and has the goal of becoming the “PC of biology labs.”
Pharmaceutical giant AstraZeneca is behind an open innovation initiative to develop next-generation medicines and technologies. According to the Alliance of Advanced Biomedical Engineering:

A cornerstone of the initiative is NiCoLA-B, a drug discovery robot at the U.K. Center for Lead Discovery, the company’s research center on the Cambridge campus. Its name is a variation of the moniker for the entire robot system: CoLAB (collaborative laboratory), which was developed by HighRes Biosolutions. The robot can test more than 300,000 compounds a day in a ballet of procedures with a central, mechanical arm as the lead dancer: the company says it is the fastest of its kind in the world. AstraZeneca explains that the robot uses sound waves to move droplets of potential drugs, billionths of liters at a time, from storage tubes into miniature wells on assay plates. Droplets of cells or biochemical solutions are added to the wells, and the robot monitors interactions for any activity that could indicate a promising drug.

One of the tremendous advantages of a system like this over human biochemists is that the robot can run for 24 hours per day, seven days a week.
The IBM system is an exciting illustration of the advantages that AI and automation technologies, especially robots, can bring to the drug discovery process. Given the times we live in, we can use all the advantages we can get. More

With drug development front page news around the world, a quieter story is spreading in automation circles: The rise of robots as indispensable lab tools in drug development. As the need for faster drug development surges, manufacturers and researchers have adopted systems that promise new levels of efficiency and accuracy.
My colleague reported on the latest example this week, an AI platform and automated drug lab from IBM, which could cut the cost and time associated with drug discovery and synthesis. The IBM system uses computational modeling and predicts chemical reactions in order to aid human chemists in the creation of new molecules. The process allows biochemists to work remotely via an automated lab.

A chemist could be sitting at home who was willing to make a molecule, and after connecting to RoboRXN for Chemistry via a web browser, they draw the molecule.
RoboRXN would then recommend optimal scientific routes and the best starting material available commercially. 
Once submitted, RoboRXN would self-program itself to “execute the process in an autonomous laboratory.” In other words, experiments could be conducted remotely with the right integration and hardware setup.

It’s part of a broader trend toward automation in the hard sciences, which, somewhat ironically given their scientific bent, have been slower than many other industries to embrace automation technology. The barrier to automation has long been a difficult combination of intense regulation, budget constraints, and a trade tradition of seasoned craftsmanship on the lab bench, all of which have slowed the adoption of lab automation.
There has also long been a labor case against automating laboratories. Grad students and interns work impossible hours for little or no pay in academic and clinical settings, where future scientists and researchers learn the ropes and much of the science underlying drug development originates. And because labs often fall under the managing jurisdiction of a highly-credentialed and invariably busy person, it can be difficult for companies to market their products to anyone with decision-making capabilities.
The coronavirus has upended that paradigm. With barriers to working in person, many research institutions are left scrambling to find a way to do hard science. Essential worker exemptions have allowed many drug manufacturers to continue working (a Gilead facility in the south bay region of Los Angeles is thrumming with activity these days; Gilead manufactures Remdesivir, which is being used in some COVID-19 treatment protocols), but the lockdowns have only accelerated the glance toward automation in the hard sciences, a trend seen in industries as disparate as fast food and grocery delivery.
That realization coincides with a decade of falling sensor prices and increasing market penetration of robots, which have brought the costs of automation technology down. That’s led to spiking interest in automating tasks like liquid-handling, which is a notoriously error-prone component of work in the hard sciences.
A Swiss company called Andrew Alliance, for example, makes liquid-handling robots for life sciences labs and has raised millions in backing. Andrew Alliance’s flagship product is a tabletop robot that fits in a footprint about the size of piece of notebook paper. Thanks to advances in collaborative automation, the robot is ready to go out of the box and is easy to operate.
“Demand for better reproducibility continues to increase,” said Otello Stampacchia, managing director and founder of Omega Funds, an early investor in Andrew Alliance.

Another company called Opentrons makes a pipetting robot for use in labs and has the goal of becoming the “PC of biology labs.”
Pharmaceutical giant AstraZeneca is behind an open innovation initiative to develop next-generation medicines and technologies. According to the Alliance of Advanced Biomedical Engineering:

A cornerstone of the initiative is NiCoLA-B, a drug discovery robot at the U.K. Center for Lead Discovery, the company’s research center on the Cambridge campus. Its name is a variation of the moniker for the entire robot system: CoLAB (collaborative laboratory), which was developed by HighRes Biosolutions. The robot can test more than 300,000 compounds a day in a ballet of procedures with a central, mechanical arm as the lead dancer: the company says it is the fastest of its kind in the world. AstraZeneca explains that the robot uses sound waves to move droplets of potential drugs, billionths of liters at a time, from storage tubes into miniature wells on assay plates. Droplets of cells or biochemical solutions are added to the wells, and the robot monitors interactions for any activity that could indicate a promising drug.

One of the tremendous advantages of a system like this over human biochemists is that the robot can run for 24 hours per day, seven days a week.
The IBM system is an exciting illustration of the advantages that AI and automation technologies, especially robots, can bring to the drug discovery process. Given the times we live in, we can use all the advantages we can get. More

As autonomous cars roll out in testbeds around the world, it should come as no surprise that other kinds of vehicles are ditching their human operators as well. The latest innovations include airplanes, and one company’s pilotless debut hints a future of pilot-less air transport.
The company, Xwing, has now performed numerous passenger-carrying autonomous take-off to landing flights in its modified Cessna 208B Grand Caravan, a first in aviation with this category of aircraft. The company has been operating in stealth as it targets regional passenger and cargo operations within a 500 mile range. 
Unmanned drones have now long been a part of the aerial landscape, but drones aren’t the only kind of self-driving aerial vehicle regulators have been dealing with. It may seem a foregone conclusion that self-driving cars are on the way, but we’ve heard less about autonomous aircraft. That’s changing. Following recent crashes related to failures in autonomous systems on-board Boeing’s 737MAX, you might expect consumer confidence to have eroded significantly. However, a recent ANSYS study found that wasn’t the case. In fact, 70% of consumers say they are ready to fly in autonomous aircraft in their lifetime. 
More recently, several companies have debuted air taxis, which promise to whisk passengers above traffic en route to their destination. 
Xwing is targeting the regional air travel market. The company has obtained a Part 135 Air Carrier certificate from the FAA and will look to start commercial cargo flight operations using its own fleet in the coming months. 
The case for autonomous air travel is actually pretty strong. The industry is dealing with a 30% reduction in qualified pilots over the last 30 years, according to the FAA. The boom in logistics and fulfillment, and especially next-day delivery, has added tremendous burdens to the air shipping ecosystem.
Xwing’s Autoflight System addresses the problem by retrofiting existing aircraft into what it calls “optionally piloted vehicles.” Remote operators who work with air traffic controllers to ensure safety throughout the flight monitor operations from the ground, centralizing human resources. According to the company, that yields 20-30 percent cost savings.
“The future of air transportation is autonomous,” said Marc Piette, CEO and founder of Xwing. “We believe the path to full autonomy begins with the air cargo market, and involves remote operators supervising fleets of unmanned aircraft.”
Xwing has been working with the FAA to obtain certification for unmanned 9,000lb Cessna 208B Grand Caravan aircraft with cargo capacity of over 4,000lbs, a type of plane used frequently for regional cargo and humanitarian missions throughout the world.  More

A Finish startup has been climbing the walls during the pandemic. At least the crops it helps grow in vertical gardens have been, including greens, berries, and vegetables in areas like the Middle East.
Vertical farming, which utilizes vertically-stacked layers of crops grown in climate-controlled facilities, utilizes significantly less water and soil than traditional agriculture. Increasingly we’re seeing examples of the concept scaling to industrial-levels, which is good news with populations booming, arable land in ever-shorter supply, and waning interest in agriculture among city-bound youth.
iFarm has figured out a smart value proposition in the still-nascent market as a developer of vertical farm management technology, essentially an operating system that utilizes tremendous volumes of sensor data to fine tune automated crop growing. The company believes it’s entering a market primed for steep growth.
“Investors can participate in the worldwide network of vertical farms and receive a rate of return well above bank deposit rates.”, says Alex Lyskovsky, co-founder and President of iFarm. “We already have a group of financial partners involved in the development of our farms, and now there is a direct opportunity for this type of investment in Finland, UK, Switzerland, Netherlands, Russia and UAE.”
One of the interesting advantages of vertical farming, particularly in a pandemic when so many professional spaces stand empty, is that it’s possible to utilize the urban environment to facilitate crop growing. By growing crops closer to city dwellers, the company can offer logistics efficiencies and unparalleled freshness. 
This at a time when traditional farming is less and less viable. Global agricultural productivity is suddenly slowing for the first time in decades. No one is quite sure why, but it’s likely a systemic problem related to the rise of monocultures and the overuse of fertilizers, which add harmful salts to soils. Farmers are also aging globally as younger generations migrate to cities. That’s largely because a productivity boom over the last century has kept food prices low, which makes farming unattractive economically. It’s a double whammy now that that productivity can no longer be taken for granted without major rethinks to the food supply chain.
Vertical farming and other smart agriculture innovations may offer realistic alternatives, and they’ve captured imaginations due to novel use of space and cutting edge technologies. iFarm’s Growtune tech platform allows growers to leverage technologies like computer vision, machine learning, and huge volumes of data data. The system can enable farming operations to spread vertical farms across distributed networks while still maintaining centralized control. And if there’s any doubt that farming has changed, the level of control is staggering. The Growtune platform can determine the plant’s weight, as well as growth deviations or pathologies, and build a system that improves crop quality and characteristics on its own. According to iFarm, the optimization will reduce labor costs for crops like strawberries, cherry tomatoes, sweet peppers, radish, and others. 
“The 2020 pandemic exposed the problems of the global food system – food supplies, sowing and harvesting were disrupted across the globe”, says Mikhail Taver, Managing Partner at Gagarin Capital. “iFarm is taking a novel approach to agriculture, offering an automated solution to grow crops close to the consumer and ensure food security. We believe that the future of the food market lies in modern technologies and are excited to support the project on its way.” More

Sensor packages are becoming ever-more-dynamic in robotics development. Robots are increasingly savvy at touching, sensing minute forces, and seeing in a variety of spectrums. I’ve written about the sense of touch in particular as a paradigm-expanding extension of the sensing toolkit.
The ability to hear has been lurking more quietly behind the scenes, in large part because the sensors required to hear are so rudimentary that they’ve been taken for granted. But most humans rely heavily on sound to orient themselves in the world, so it stands to reason that automation controls system should as well.
Now researchers at CMU’s Robotics Institute have found that sounds may actually be used to enable a robot to better tell one objects from another. After all, the objects have different physical properties that produce different sounds when they’re handled, clanged against something, or used in the field.
“A lot of preliminary work in other fields indicated that sound could be useful, but it wasn’t clear how useful it would be in robotics,” explains Lerrel Pinto, who recently earned his Ph.D. in robotics at CMU and will join the faculty of New York University this fall. 
Researchers at the Robotics Institute created a large dataset of video and audio recordings of everday objects as they slid or rolled around a tray and crashed into its sides. The way this database was populated is a mini story within the story given its savvy deployment of automation. The researchers used a robot named Sawyer from defunct developer Rethink Robotics. Sawyer held a tray and each object — a tennis ball, for example, or a toy block — was placed on the tray. Sawyer then spent hours moving the tray in random directions while cameras and microphones recorded everything. In all the researchers captured a dataset of about 15,000 interactions, which they’ve since released. 
[embedded content]
Drawing on parallel research into the use of sound to help robots estimate the trajectories and movement of object, the researchers reinforced the usefulness of sound for robots, finding that robots equipped with the insights gleaned from the dataset managed successful classifications of objects about three-quarters of the time based on sound alone.
“I think what was really exciting was that when it failed, it would fail on things you expect it to fail on,” he said. For instance, a robot couldn’t use sound to tell the difference between a red block or a green block. “But if it was a different object, such as a block versus a cup, it could figure that out.”
Interestingly, the Department of Defense’s research grant arm, DARPA, supported the research, along with the Office of Naval Research, both big investors in automation technologies and research.   More

MaxPixel’s contributors
Smart traffic lights may make life easier for drivers in Phoenix. Here in LA, where traffic is now picking up again despite the ongoing pandemic, I can safely say we’re monitoring the situation closely.
The new effort is the brainchild (naturally) of a California-based company called NoTraffic. The rollout, which begins today, will take place a few key intersections in order to improve traffic flow and reduce vehicle and pedestrian delays by taking the system off of a timer-based model and coordinate the lights based on actual demand. In some deployments, the company has seen up to 40 percent reduction in vehicle delay time.
“We are grateful to partner with the City of Phoenix, the fifth largest city in the United States,” says Tal Kreisler, CEO of NoTraffic. “I believe we will see impactful tech initiatives moving front and center, playing a pivotal role in how the world emerges from COVID-19 and the economic recession that accompanies it.”
One of the interesting benefits of the traffic grid management system is that it can coordinate what’s called emergency vehicle preemption to give first-responders the clearest path through busy commuter corridors. Leveraging AI and connected vehicle technology (V2X) that distinguishes between cars, bikes, pedestrians, buses, emergency vehicles, and commercial fleets, the NoTraffic platform tracks road assets as they approach an intersection and calculate “in real time the most optimal service for the intersection and autonomously changing the lights accordingly.”  The system takes into  account safety considerations like vehicles’ blind spots.
“We are now seeing the convergence of technology-enabled automobiles and traffic management systems working together to move vehicles more effectively through busy corridors,” Phoenix Street Transportation Director Kini Knudson said. “The opportunity to collaborate and test this new technology with regional and global partners is very exciting for Phoenix.”
The Phoenix deployment, which will test run the system for a potential larger rollout, is part of the Maricopa Association of Governments (MAG) emerging technologies initiative, which tests new technologies for viability before large-scale investments are made.”As the regional transportation planning agency that serves all residents in the region, MAG continually evaluates smart mobility systems and works closely with jurisdictions that champion innovation,” said MAG Executive Director Eric Anderson. “It’s important to safely deploy technologies that are ready in the real world. Working together with our university partners, we look forward to seeing performance of the NoTraffic platform in busy commuter and retail corridors.” More

A company that seemingly came out of nowhere in 2018 and has been leading the pack in self-driving services ever since is about to corner its home market, the globally strategic city of Shanghai. Since April, AutoX has been running a testbed of its autonomous taxi service, and now the rollout will cement its headstart.
AutoX recently made news in California for winning a coveted permit to test its driverless cars without drivers, just the third company to be awarded the permit. Meanwhile, in China, the company, which is helmed by a former Princeton professor known affectionately as Professor X, has made start strategic partnerships, including with Alibaba. AutoX’s autonomous taxi pilot program allowed users to hail the taxis with Alibaba’s map product,  AutoNavi.
The so-called RoboTaxi ride-hailing service is now open generally to the public for the first time, and the AutoNavi partnership will carry over. Additionally, AutoX has struck up a strategic partnership with a major taxi fleet operations company called Letzgo, illustrating a savvy propensity for operations expansion through alliance. The partnership will allow AutoX RoboTaxis to be hailed through Letzgo’s smartphone app, on top of Alibaba’s AutoNavi app, in the coming weeks. 
On its surface Letzgo may seem like a peculiar partner. The company has 16,000 human-operated vehicles. But the writing is on the wall for rideshare services. Lacking the ability to develop its own self-driving development, ala Uber, Letzgo is teaming up with AutoX and vowing to retrain staff to operate RoboTaxis. It’s hard to believe the longterm model will leave room for the same scale of human-led workforce, however. 
The appeal of self-driving taxis is even greater as the pandemic, largely under control in China, rages elsewhere in the world. AutoX says it has taken special measures, including a voice recognition system instead of touchscreens. AutoX is currently testing its vehicles in Shenzhen, Wuhan, Wuhu, and several other cities around the world.  More

One of the most interesting technology stories of the past few years has centered on a decidedly legacy industry: Construction. There’s increasing competition between technology developers to bring new efficiencies to the job site, and it’s coming from firms working in areas like robotics and artificial intelligence.
I was reminded of this when I learned that a company called Buildots is pulling down $16M for their AI construction solution which is based around hardhat-mounted 360° cameras. That’s a relatively benign tech upgrade, perhaps, but it’s part of a larger trend: Much of the tech development in construction is focused on augmenting humans with technology, pointing to a future of hybrid man-machine workers.
Some of the more eye-catching technologies coming online now include wearable robots. A company called Sarcos, for example, makes a robotic exoskeleton that enables humans to lift and manipulate incredibly heavy objects with ease. While these suits aren’t yet common on job sites, it’s easy to see what’s coming.
The reason for the rush to make human construction workers stronger, faster, and smarter is related to the disquieting fact that productivity in construction has actually fallen in half since the 1960s. The sector has not kept pace with innovation, and as I’ve written, the diesel-powered hydraulic machines you’ll find on most construction sites today remain essentially unchanged from those rolling around 100 years ago. 
As a result, there are massive inefficiencies in the industry. According to KPMG’s Global Construction Survey, just 25% of projects came within 10 percent of their original deadlines. When it comes to megaprojects, like large infrastructure projects, McKinsey found that 98% are delayed or over budget. 77% are more than 40% behind schedule.
In contrast to Sarcos’ robot, the Buildots camera is a more incremental tech add-on. The idea is that workers or project managers wear cameras to capture every detail during daily site walkthroughs. It brings to mind a system developed by competitor OpenSpace, which came out of stealth a couple years back. Using AI, the captured information is stitched together in a useful way. OpenSpace focuses on project overruns, promising its solution helps keep building on schedule. Buildots’ platform uses the captured imagery and compares it to the building’s design plans, checking for the slightest deviation or omission. “Think a missing electricity point or an AC vent 3 inches to the left,” a company rep told me. 
This could help stem a big inefficiency problem in the sector, particularly as housing crunches in places like Los Angeles, where I live, intensify. Right now, project managers record mistakes and anomalies manually and, depending on the size of the project, they may be expected to keep track of 100,000+ tasks. McKinsey recently found that construction inefficiency is costing the industry $1.6 trillion a year. That represents a massive opportunity for technology firms, which are now racing into the space.
Buildots, a UK and Israel-based company, was founded in 2018 by three graduates from ‘Talpiot’, the creative engine of the Israeli Defense Forces and its most prestigious unit. It counts among its clients two of the ten largest construction companies in Europe, as well as the largest Israeli construction firm. Riding the new cash infusion, Buildots has plans to make their product available to the US market in the coming year. More