Einride demonstrates a single operator controlling multiple of its driverless cargo vehicles

Autonomous electric transportation startup Einride has taken a key step in its mission to deploy autonomous cargo pods on roads for commercial operations. The Swedish startup demonstrated its technology in use with one person remotely operating two pods at once, which is a fundamental part of their vision of multiple pods ultimately being overseen by one person essentially operating as a traffic controller.

The demonstration saw an operator oversee and remotely control the two driverless pods using a steering wheel controller and a surround view display using a number of monitors. The system demo shows how a pod can request that an operator take over manual control if it encounters an issue it can’t address via its onboard automated driving computer.

It’s a clever and practical way to bridge the gap between manually driven vehicles and fully autonomous transportation, while still changing the economics of fleet logistics. With a one-to-many model, Einride would be able to offer trucking companies big advantages in terms of costs and efficiencies, increasing the number of miles that can be driven without boosting headcount requirements. Plus, the electric drivetrains of the vehicles will add up to big fuel and ecological advantages when it comes to day-to-day operations.

Einride also says that its platform has the potential to change the dynamics of the profession of trucker, since it can provide comfortable, remote operations centers that replace long weeks on the road away form home. This could open up the industry to more potential employees and recruits, which is a crucial need since trucking has typically required more new drivers than the market could supply in the U.S. over the pas few years.

Einride’s demonstration included complex maneuvers including parking and pulling out from a busy transportation hub, and shows in practice the potential of their tech. The company announced a commercial trial with Coca-Cola’s official European bottling and distribution partner at the end of last year, and is continuing to work towards broad commercialization.

Google research makes for an effortless robotic dog trot

As capable as robots are, the original animals after which they tend to be designed are always much, much better. That’s partly because it’s difficult to learn how to walk like a dog directly from a dog — but this research from Google’s AI labs make it considerably easier.

The goal of this research, a collaboration with UC Berkeley, was to find a way to efficiently and automatically transfer “agile behaviors” like a light-footed trot or spin from their source (a good dog) to a quadrupedal robot. This sort of thing has been done before, but as the researchers’ blog post points out, the established training process can often “require a great deal of expert insight, and often involves a lengthy reward tuning process for each desired skill.”

That doesn’t scale well, naturally, but that manual tuning is necessary to make sure the animal’s movements are approximated well by the robot. Even a very doglike robot isn’t actually a dog, and the way a dog moves may not be exactly the way the robot should, leading the latter to fall down, lock up, or otherwise fail.

The Google AI project addresses this by adding a bit of controlled chaos to the normal order of things. Ordinarily, the dog’s motions would be captured and key points like feet and joints would be carefully tracked. These points would be approximated to the robot’s in a digital simulation where a virtual version of the robot attempts to imitate the motions of the dog with its own, learning as it goes.

So far, so good, but the real problem comes when you try to use the results of that simulation to control an actual robot. The real world isn’t a 2D plane with idealized friction rules and all that. Unfortunately, that means that uncorrected simulation-based gaits tend to walk a robot right into the ground.

To prevent this, the researchers introduced an element of randomness to the physical parameters used in the simulation, making the virtual robot weigh more, or have weaker motors, or experience greater friction with the ground. This made the machine learning model describing how to walk have to account for all kinds of small variances and the complications they create down the line — and how to counteract them.

Learning to accommodate for that randomness made the learned walking method far more robust in the real world, leading to a passable imitation of the target dog walk, and even more complicated moves like turns and spins, without any manual intervention and only little extra virtual training.

Naturally manual tweaking could still be added to the mix if desired, but as it stands this is a large improvement over what could previously be done totally automatically.

In another research project described in the same post, another set of researchers describe a robot teaching itself to walk on its own, but imbued with the intelligence to avoid walking outside its designated area and to pick itself up when it falls. With those basic skills baked in, the robot was able to amble around its training area continuously with no human intervention, learning quite respectable locomotion skills.

The paper on learning agile behaviors from animals can be read here, while the one on robots learning to walk on their own (a collaboration with Berkeley and the Georgia Institute of Technology) is here.

OrbitFab secures National Science Foundation funding to propel its satellite refueling tech to space

On-orbit satellite refueling technology is closer than ever to a practical reality, which could help immensely with the cost and sustainability of orbital businesses. Startup OrbitFab, a 2019 TechCrunch Battlefield finalist, is one of the companies working to make orbital refueling a reality, and it just secured a new contract from the National Science Foundation’s early stage deep tech R&D initiative America’s Seed Fund to further its goals.

The contract is specifically for development of a solution that provides rendezvous and docking capabilities in space, managing the end-to-end process of connecting two spacecraft and transferring fuel from one to the other. OrbitFab unveiled its connector hardware for making this possible last October at Disrupt, which it now refers to as its Rapidly attachable Fluid Transfer Interface (RAFTI). The RAFTI is designed as a replacement for existing valves used in satellites for fueling and draining propellant from spacecraft, but would seek to establish a new standard that provides easy interoperability both with ground fueling, and with in-space refueling (or fuel transfer from one satellite to another, depending on what’s needed).

Already, OrbitFab has managed to fly twice to the International Space Station (ISS), and last year it became the first ever private company to supply the orbital lab with water. It’s not resting on its laurels, and this new contract will help it prepare a technology demonstration of the docking process it’s RAFTI facilitates in its own test facilities this summer.

Longer-term, this is just phase one of a multi-par funding agreement with the NSF. Phase one includes $250,000 to make that first demo, and then ultimately that will lead to an inaugural trial of a fuel sale operation in space, which OrbitFab CMO Jeremy Schiel says should happen “within two years.”

“This will involve 2 satellites, our tanker, and a customer satellite, in a low LEO [low Earth orbit] docking, exchanging fuel, and decoupling, and repeating this process as many times as we can to demonstrate our capability,” he wrote via email.

There have been a number of technical projects and demonstrations around orbital refueling, and some of the largest companies in the industry are working on the challenge. But OrbitFab’s approach is aiming for simplicity, and ease of execution, along with a common standard that can be leveraged across a wide range of satellites large and small, from a range of companies. Already, OrbitFab says it’s working with a group of 30 different campaigns and organizations on making RAFTI a broadly adopted interface.

If successful, OrbitFab could underpin a future orbital commercial operating environment in which fuel isn’t nearly as much a concern when it comes to launch costs, with on-orbit roving gas stations addressing demand for spacecraft once they reach space, and paying a price for propellant that’s defrayed by common, bulk shipments instead of broken up piecemeal.