Soft robots can be used in various spheres, such as agriculture, medicine, and defense. However, their complex physics means that they are hard to control. Current simulation testbeds are insufficient for taking the full advantage of elasticity.
A recent paper on arXiv.org proposes Elastica, a simulation environment tailored to soft robot context. It tries to fill the gap between conventional rigid body solvers, which are incapable to model complex continuum mechanics, and high-fidelity finite elements methods, which are mathematically cumbersome. Elastica can be used to simulate assemblies of soft, slender, and compliant rods and interface with major reinforcement learning packages. It is shown how most reinforcement learning models can learn to control a soft arm and to complete successively challenging tasks, like 3D tracking of a target, or maneuvering between structured and unstructured obstacles.
Soft robots are notoriously hard to control. This is partly due to the scarcity of models able to capture their complex continuum mechanics, resulting in a lack of control methodologies that take full advantage of body compliance. Currently available simulation methods are either too computational demanding or overly simplistic in their physical assumptions, leading to a paucity of available simulation resources for developing such control schemes. To address this, we introduce Elastica, a free, open-source simulation environment for soft, slender rods that can bend, twist, shear and stretch. We demonstrate how Elastica can be coupled with five state-of-the-art reinforcement learning algorithms to successfully control a soft, compliant robotic arm and complete increasingly challenging tasks.
Fast and Robust Bio-inspired Teach and Repeat Navigation
The navigation of robots is a demanding task. Luckily, we can rely on biological systems, such as ants, which can navigate with limited vision and computing power. A recent study suggests a teach and repeat navigation system for repeated route following.
It is based on wheel odometry, with vision providing a periodic correction signal. This technique can be applied to small low-cost robots, which usually have wheel odometry sensors and a monocular camera but do not have stereo vision or LiDAR sensors. The rate of visual correction can be changed accordingly to available computation resources.
Example of a navigation Doppler lidar instrument. Credits: NASA
The approach was verified in indoor and outdoor trials at different times of the day and in varying weather conditions. It can be used for new robotic systems with minimal tuning. The method is robust to odometry errors and can work with low-resolution images.
Fully autonomous mobile robots have a multitude of potential applications, but guaranteeing robust navigation performance remains an open research problem. For many tasks such as repeated infrastructure inspection, item delivery or inventory transport, a route repeating capability rather than full navigation stack can be sufficient and offers potential practical advantages. Previous teach and repeat research has achieved high performance in difficult conditions generally by using sophisticated, often expensive sensors, and has often had high computational requirements. Biological systems, such as small animals and insects like seeing ants, offer a proof of concept that robust and generalisable navigation can be achieved with extremely limited visual systems and computing power. In this work we create a novel asynchronous formulation for teach and repeat navigation that fully utilises odometry information, paired with a correction signal driven by much more computationally lightweight visual processing than is typically required. This correction signal is also decoupled from the robot’s motor control, allowing its rate to be modulated by the available computing capacity. We evaluate this approach with extensive experimentation on two different robotic platforms, the Consequential Robotics Miro and the Clearpath Jackal robots, across navigation trials totalling more than 6000 metres in a range of challenging indoor and outdoor environments. Our approach is more robust and requires significantly less compute than the state-of-the-art. It is also capable of intervention-free — no parameter changes required — cross-platform generalisation, learning to navigate a route on one robot and repeating that route on a different type of robot with different camera.
Facebook Demands Academics Disable Ad-Targeting Data Tool
Academics, journalists, and First Amendment lawyers are rallying behind New York University researchers in a showdown with Facebook over its demand that they halt the collection of data showing who is being micro-targeted by political ads on the world’s dominant social media platform.
The researchers say the disputed tool is vital to understanding how Facebook has been used as a conduit for disinformation and manipulation.
In an October 16 letter to the researchers, a Facebook executive demanded they disable a special plug-in for Chrome and Firefox browsers used by 6,500 volunteers across the United States and delete the data obtained. The plug-in lets researchers see which ads are shown to each volunteer; Facebook lets advertisers tailor ads based on specific demographics that go far beyond race, age, gender and political preference.
The executive, Allison Hendrix, said the tool violates Facebook rules prohibiting automated bulk collection of data from the site. Her letter threatened “additional enforcement action” if the takedown was not effected by Nov. 30.
Company spokesman Joe Osborne said in an emailed statement Saturday that Facebook “informed NYU months ago that moving forward with a project to scrape people’s Facebook information would violate our terms.” The company has long claimed protecting user privacy is its main concern, though NYU researchers say their tool is programmed so the data collected from participating volunteers is anonymous.
The outcry over Facebook’s threat was immediate after The Wall Street Journal first reported the news Friday considering the valuable insights the “Ad Observer” tool provides. It has been used since its September launch by local reporters from Wisconsin to Utah to Florida to write about the November 3 presidential election.
“That Facebook is trying to shut down a tool crucial to exposing disinformation in the run up to one of the most consequential elections in US history is alarming,” said Ramya Krishnan, an attorney with the Knight First Amendment Institute at Columbia University, which is representing the researchers. “The public has a right to know what political ads are being run and how they are being targeted. Facebook shouldn’t be allowed to be the gatekeeper to information necessary to safeguard our democracy. “
“The NYU Ad Observatory is the only window researchers have to see microtargeting information about political ads on Facebook,” Julia Angwin, editor of the data-centric investigative tech news website The Markup, tweeted in disappointment.
The tool lets researchers see how some Facebook advertisers use data gathered by the company to profile citizens “and send them misinformation about candidates and policies that are designed to influence or even suppress their vote,” Damon McCoy, an NYU professor involved in the project, said in a statement.
After an uproar over its lack of transparency on political ads Facebook ran ahead of the 2016 election, a sharp contrast to how ads are regulated on traditional media, the company created an ad archive that includes details such as who paid for an ad and when it ran. But Facebook does not share information about who gets served the ad.
The company has resisted allowing researchers access to the platform, where right-wing content has consistently been trending in recent weeks. Last year, more than 200 researchers signed a letter to Facebook calling on it to lift restrictions on public-interest research and journalism that would permit automated digital collection of data from the platform.
Because growth has been slow, many businesses may still be unaware of the advantages offered by products like Amazon WorkSpaces. Read on to learn why this new way of deploying workstations may be worth the investment, and the way of the future.
What is Amazon WorkSpaces?
Amazon WorkSpaces is Amazon’s entry into the DaaS field and gives businesses the ability to create persistent, virtual, and cloud-hosted Windows and Linux workstations. Like other DaaS solutions, Amazon WorkSpaces is designed to scale up as businesses grow and new workstations are needed.
Amazon describes WorkSpaces as “an easy way to provide a secure, managed, cloud-based virtual desktop experience to your end-users,” and said that its cloud-native, fully managed design means IT teams “don’t have to worry about procuring, deploying, and managing a complex environment.”
Amazon WorkSpaces virtual machines (VMs) can be deployed in both Linux and Windows formats and are designed to fill the role of basic end-user workstations. To fill various workstation roles, a number of different bundles are available that allocate more or less CPU power, GBs of RAM, GPUs, video memory, SSD root and user storage, and software.
IT teams can use default Windows/Linux images or create their own custom ones to ensure business-essential apps are installed on new VMs; Amazon also offers Amazon WorkSpaces Application Manager (Amazon WAM) for deploying and managing additional applications on WorkSpaces VMs. Amazon WAM packages applications into containers and makes them appear on user WorkSpaces instances as if they were locally installed, while administrators maintain the ability to manage them as containerized apps, which eliminates the need for time-consuming update, deployment, and retirement cycles.
In terms of the security of WorkSpaces, Amazon said it uses encryption strong enough to meet HIPAA and PCI compliance through its PC-over-IP protocol, which encrypts traffic and doesn’t send any data to, or store anything sensitive on, end-user devices. User access can also be restricted via IP address, device type, or with digital certificates, and the entire system is integrated with AWS Key Management Service for encrypting storage volumes.
Credentialing for Amazon WorkSpaces access can be handled in a number of ways, including using pre-existing on-premise Microsoft Active Directory installations linked to AWS Directory Service. In that instance, users will be able to log in to their WorkSpaces instance using their existing credentials, and IT teams can apply group policy settings as usual, deploy software with existing tools, and use existing RADIUS servers to enable multifactor authentication.
On the user end, accessing an Amazon WorkSpaces VM can be done on pretty much any device imaginable: Downloadable clients exist for Windows 10, macOS, iPad, Fire tablets, Android, ChromeOS, and Linux. There’s also a web client for those who don’t want to install software, too. When a new Amazon WorkSpaces instance is provisioned for a user, they receive an email with a link to download the client (or use the web client), verify their identity, and that’s it–they’re all set to work on a virtual, managed, and secured desktop from wherever they are.
What are use cases for Amazon WorkSpaces?
It’s not hard to see use cases for Amazon WorksSpaces almost immediately, especially in the post-COVID-19 age. The fact that remote workers can be issued an encrypted, managed, and persistent virtual machine to work from anywhere they are can be transformative.
Amazon mentions several basic use cases on the WorkSpaces website that can be applied to any number of industries and organizations:
Quickly provision secure desktops for remote, mobile, and contract employees;
Since it meets those various certifications, WorkSpaces can be used for online payments, storing customer data securely in the cloud, US government and Department of Defense computing, healthcare needs, and more.
What are DaaS alternatives, and how much does Amazon WorkSpaces cost?
In an article about top DaaS providers, TechRepublic editor in chief Bill Detwiler gave a rundown of the biggest names in DaaS and reasons why each one would be a good choice for different types of organizations. Aside from Amazon WorkSpaces, nine DaaS providers are mentioned in his article: Citrix Managed Desktops, Cloudalize DaaS, dinCloud dinWorkspace, Evolve IP, itopia Cloud Automation Stack, Microsoft Windows Virtual Desktop, MTM Technologies AnywhereApp, and VMware Horizon Cloud.
Of the DaaS options listed above, Amazon WorkSpaces may be the most affordable choice, or at the very least the most flexible: Many DaaS providers require a minimum number of seats, longer-term commitments, or their prices are simply higher for pay-as-you-go DaaS instances.
Amazon WorkSpaces, on the other hand, has no minimum term, and prices for a single instance start at $7.25 per month plus 17 cents per hour.
Like any kind of new technological investment, it’s important to look at each vendor and compare features and pricing to determine what best suits your needs. Amazon WorkSpaces offers a lot of good features, but if you’re not already an AWS customer, it may be better to look at another provider, like MTM Technologies AnywhereApp, which is compatible with AWS, Azure, CenturyLink, and Oracle public clouds. Organizations already operating on Azure would be better suited to using Microsoft’s own Windows Virtual Desktop, and Google Cloud Platform users only have one option from Detwiler’s list: itopia Cloud Automation Stack.
In the intro to this cheat sheet, I mentioned the battle between virtual desktop infrastructure (VDI) and DaaS, two completely different technologies that perform the same task: Creating virtual desktops for users to work on in place of their local machine.
VDI is not a cloud-based product–it’s the traditional virtualization product installed and operated from an organization’s data center. DaaS is VDI running on a cloud provider’s hardware. In the case of Amazon WorkSpaces, that hardware is owned and operated by Amazon Web Services.
Several immediate reasons to choose Amazon WorkSpaces over a locally-installed VDI product come to mind immediately, particularly the elimination of in-house hardware dedicated to running VDIs. As organizations grow, VDI hardware may need to be scaled, it can get old and need replacing or repair, data center disasters can occur and cause downtime, and increasingly distributed workforces make managing VDIs difficult for IT teams.
DaaS solutions like Amazon WorkSpaces are just one more data center resource that can be foisted off on the cloud, freeing up space in the server room, eliminating hardware management needs, and ensuring more uptime. As COVID-19 transforms the work world into a largely remote one, DaaS solutions like Amazon WorkSpaces are becoming even more valuable for forward-looking companies that are considering permanently eliminating physical office spaces. In that kind of future, a VDI server is just one more piece of equipment that won’t have anywhere to go but the cloud.
How can I start using Amazon WorkSpaces?
It’s not too difficult to start with Amazon WorkSpaces: Interested parties will need an AWS account, basic knowledge of the AWS console in order to find the WorkSpaces section, and time to customize a WorkSpaces instance for deployment, all of which can be done in the AWS console.
Amazon provides a helpful getting started with WorkSpaces guide on its website, which includes basic steps, starter projects, and best practices documentation.
Organizations looking to enter the Amazon WorkSpaces DaaS sphere can do so free of charge using the WorkSpaces free tier, which provides two standard bundle WorkSpaces instances for up to 40 combined usage hours per month for two months. After that, the two instances are billed per hour. Educational pricing is also available for qualified Windows educational users, who will receive a discount of $3.52 per user per month, and a 3 cents per user per hour discount on hourly usage.
Pricing varies greatly based on the type of machine, its specifications, and whether application bundles are needed. For pricing details, check out Amazon’s WorkSpaces pricing page.
Cloud and Everything as a Service Newsletter
This is your go-to resource for XaaS, AWS, Microsoft Azure, Google Cloud Platform, cloud engineering jobs, and cloud security news and tips. Delivered Mondays