The chip, which is specifically made for autonomous applications, will provide the robust, compact, energy efficient and auto-grade compute power necessary to accelerate TuSimple’s race to get its self-driving trucks to market. The CEO and president of TuSimple, Cheng Lu said:

“Without an auto-grade ADC, it’s not possible for any OEM or autonomous technology player to have scaled production of AV trucks on the road,” Lu told TechCrunch. “Today we use prototype systems that are costly, and not as reliable. It works for a small batch of vehicles, but would not work when you manufacture integrated AV trucks at the OEM at scale.”

TuSimple plans to build semi trucks specifically designed for autonomous applications by 2024. The project is going to happen in collaboration with Navistar and the Traton Group, Volkswagen AG’s heavy truck business.

The Drive Orin SoC delivers 254 trillion operations per second of performance, including perception, planning and actuation functions. Along with this most critical hardware component of the ADC, Nvidia brings to the partnership a good understanding of the design and placement of chips into a computational box to make them work most efficiently, says Lu.

The company provides requirements like specific computational needs, sensors and power usage, because TuSimple’s software is embedded into the ADC.

“Elliptic Labs’ partnership with Xiaomi is solid and we’re pleased that they continue to use us in more smartphones as they have grown into the second-largest smartphone OEM in the world. The AI Virtual Smart Sensor Platform allows manufacturers to introduce new features, while omitting hardware-based sensors, reducing cost and eliminating sourcing risk. We are confident that our AI Virtual Smart Sensor Platform will continue to provide the economies of scale, performance, and value that innovative partners like Xiaomi require to continue such growth.”

INNER BEAUTY is a registered trademark of Elliptic Labs.

All other trademarks or service markets are the responsibility of their respective organizations.

Sheeva.AI’s API platform serves as a programmatic interface to monitor and utilize real-time vehicle location to automate payments, trigger driver-centric contextual events, and streamline consumption of vehicle services such as fueling, EV charging, parking, tolls, and curbside retail. Taking a software-first and open API philosophy, Sheeva.AI created a technology to support any mobility related service through its open API platform.

The core differentiator of Sheeva.AI’s technology is its patented location algorithm, enabling the platform to determine the vehicle’s real-time geolocation precisely enough to identify the car’s location down to the gas pump, parking space, EV charging station, toll or drive-thru lane, etc.

Evgeny Klochikhin, CEO and co-Founder of Sheeva.AI, noted:

“The mobile revolution only happened after open easy access to location-based services (aka GPS in every phone) became ubiquitous. Every car by 2025 will have the same capability:

Pervasive connectivity,

Embedded GPS, and

API-based Payments, meaning no need for complex architectures.

We at Sheeva.AI are the first to enable these services for Chipsets, OEMs, and Merchants via RESTful APIs. We deliver our solution with little to no architecture changes on behalf of our partners’ IT stacks – more will be revealed when we officially exit stealth next month.“

This level of accurate location enables the last-mile automation of services for the vehicle. Applications for this open API include curbside retail pick-up, EV Charging, paying tolls, even getting your favorite drive-thru food.

Founded by experts in robotics, Embark claims to be the longest-running self-driving truck program in the U.S., launched in 2016, when self-driving technology wasn’t so popular in the trucking industry. The company’s strategy was to focus exclusively on autonomous software and associated services, which would be universally compatible with most trucks.

So far, the company’s product range includes the Embark Driver, an advanced automated driving software that was specifically developed for trucks, the Embark Universal Interface that is used to integrate autonomous technology onto trucks, a fleet monitoring and remote-assist system.

The startup recently announced that it will start using the Nvidia Drive autonomous vehicle compute platform to power EUI and its Driver software. According to Ajith Dasari, Head of Hardware Platform at Embark, the high performance and safety standards that are required by the autonomous software and interface, can only be met with the help of “an enormous amount of computing power in our trucks”.

Known as the first scalable AI platform that combines hardware and software for the production of automated vehicles (including deep learning, surround vision and advanced sensors), Nvidia Drive was a natural choice.

The flexible EUI was specifically designed to allow carriers to select their preferred OEM (Original Equipment Manufacturer) when installing the self-driving software. At the moment, the EUI is compatible with four major OEMs – Freightliner, Navistar International, PACCAR, and Volvo – which bring most of the Class-8 trucks on U.S. roads today.

The partnership with Nvidia represents another significant step for Embark. A couple of months ago, the startup announced that it became a publicly listed company, after signing a definitive business combination agreement with Northern Genesis Acquisition Corp. II. Also, its Board of Directors was joined by former Secretary of Transportation, Elaine Chao.

So what exactly did Google add to Android 12 Beta 3? According to the official blog post, notable new features in Beta 3 include a faster display auto-rotation and scrollable screenshots.

With scrolling screenshots in Android 12, you can capture a screenshot of content that’s scrollable. This means you won’t have to capture screenshots one-by-one and then stitch them together manually. Google has been working on this feature since last year’s Android 11 developer preview, but it didn’t make the cut for that release since the team had limited resources (due to COVID-19) and wanted to implement it the right way. However, Google quietly resumed development on the feature, and we caught a glimpse of it in the first Android 12 developer preview. The scrolling screenshots implementation was hidden from subsequent developer previews and beta releases, but we saw a near-final version of the feature a few months ago thanks to a leaked internal build.

As we suspected, the way scrolling screenshots work in Android 12 is based on Views rather than the image stitching found in OEM implementations. When users capture a screenshot of content that’s scrollable, a new “capture more” button is shown that, when tapped, extends the screenshot to show the whole screen that the user can then adjust to crop.

Google says that scrolling screenshots will work out-of-the-box for most apps, provided the app uses a standard View-based UI. If an app or UI toolkit doesn’t use a View-based UI or uses a heavily customized UI, then it’ll need to implement the new ScrollCapture API to tell the system how to capture the screen. Google is working on making its scrolling screenshots implementation work in more cases, such as for scrolling ListViews (coming in Beta 4) and WebViews. Support for the latter may be coming natively to Google Chrome, but it’ll be nice to see a generic implementation that’ll work in any app that uses a WebView.

Better, faster autorotate

Android 12 is finally improving upon one of the most basic features of the OS: autorotation. Starting in Beta 3, Android’s auto-rotate feature has been enhanced with face detection, which uses the front-facing camera to more accurately detect when the screen should be rotated. This should hopefully lead to fewer unintended screen rotations when you’re lying down on a couch or in bed. Images captured by Android’s enhanced autorotation feature are processed on-device within Android 12’s Private Compute Core, so they never leave the device.

Google says that this face-based autorotation feature is available in Beta 3 on the Pixel 4 and later Pixel devices, but all devices running Android 12 can take advantage of the other improvements made to screen rotation. These improvements include optimizations to the animation and redrawing of the screen as well as an added ML-driven gesture detection algorithm, reducing the latency of the base auto-rotate feature by 25%.

Face-based autorotate was one of the earliest features rumored to be coming to Android 12, and we confirmed its existence in early developer preview builds. Based on our analysis, it seems that Google’s face-based autorotation feature is handled by the Device Personalization Services app, so it may be possible to enable it on other Pixel devices.

Douglas Johnson‑Poensgen

Hello. My name is Douglas Johnson‑Poensgen and I am the co-founder and chief executive at Circulоr.

Circulor business is traceability in supply chains. We help our customers to prove and improve their ESG performance. ESG stands behind environmental social and governance that includes, for example, demonstrating sourcing critical raw materials like example cobalt used in the lithium-ion battery supply chain, which has concerns around child labour, or other materials for example nickel which have concerns around environmental damage.

So we are actually using the flow of materials, specific bits of ESG data including embedded carbon in order to digitise the supply chain and support procurement professionals downstream brands from using procurement as a tool for driving greater sustainability, for example.

What challenges do your company address in terms of the raw materials supply chain?

We support customers to demonstrate responsible sourcing, whether that is for raw materials that cause environmental damage or association with environmental damage that will potentially be associated with child or forced labour of the supply chain. An example of that might be to trace cobalt from mines all the way through EV (electric vehicles) batteries to demonstrate that none of that cobalt has come from artisanal mine sites where there is a risk of child labour. For example geographical places like the Democratic Republic of Congo.

How’d you managed to provide the right balance of visibility and control? 

When you’re looking to manage risk within a supply chain and of course you’re digitising what’s going on within the supply chain, there’s a concern around not creating too much transparency because of course commercial confidentiality plays an important role within these supply chains. There’s a difference between traceability, which helps manage risks for example of where something has come from or to help, for example, understand the amount of embedded carbon that is being added by each participant in a supply chain, that is not the same as full transparency total disclosure of absolutely everything within that supply chain, including for example, who’s buying from whom in what quantity and what price. So there’s a difference between traceability and transparency.

What is unique about your approach? 

We were the first to be able to demonstrate the ability to track raw materials from mines all the way through the supply chain to end consumption, where those materials change between the source, multiple times on their journey through the supply chain to the point at which they get incorporated within a product. I mentioned cobalt earlier. I’m going to stick with that just so I don’t use too many analogies. Cobalt is obviously miners, or that all goes into a smelter then goes into fine refining it’s amalgamated with other materials, and eventually, that is used to manufacture the cathode, which is of course, alongside the anode, the key elements of a lithium-ion battery. That material has undergone multiple steps of transformation on his journey, and so you have to be able to connect the input to the output of each one of those industrial processes reliably multiple times through the supply chain in order to be able to demonstrate traceability.

Reuters reported that you know the first example of doing this with a compliment or called tantalum in 2018. We’re now working with a whole variety of materials, not just in the battery space but also with things like the chemical recycling of plastic waste. That at the moment is what makes our approach unique, the ability to track through industrial processes.

What’s the role of the different blockchain technologies and how do you integrate them into your business?

Well, our platform is built on hyper ledger fabric, which of course is an open-source blockchain technology layer. But blockchain itself doesn’t solve this problem of improved visibility into a supply chain in order to manage risk. You have to combine it with other technologies. In our case, that is a combination of IoT, traditional databases where a lot of the business logic is held blockchain and also machine learning, used for example for identifying anomalies in the chain of custody within the supply chain to identify places where audit action should be focused. No digital technology today can completely replace any physical due diligence, for example, that you might do at a mine site, or as a participant within the supply chain.

How is Digital Technologies transforming the supply chain?

10 years ago, the ability to digitise the supply chain was far more limited than it is today. I mean we’ve had EDI systems, for example, we’ve had, procured pay systems for some time, and within the developed world, there is a significant number of track and trace type solutions available for managing logistics. What has not been possible is to digitise the supply chain to tear, N, all the way through the supply chain to actually see what is going on within the supply chain. I think over the last 12 months of COVID, we’ve all come to recognise the relative fragility of our approach to supply chain management and the need to better understand what goes on all the way through the supply chain to the source of critical raw materials. Today, car manufacturers have for example suffered from a global shortage of semiconductors, some car manufacturers have actually had to stop production of some models because of it. And the reason for that is the inability to match demand forecasting, with a true understanding of supply chain risk, all the way back to the source of raw materials.

And I think that’s how digital technologies are starting to impact the supply chain. 10 years from now I suspect supply chains will be well digitised and these sorts of tools such as the ones we produce will have become relatively ubiquitous.

How do you manage to find good tech professionals and how do you cope with the shortage of staff within the tech industry?

We’re quite lucky because we’re a business with a social purpose. In this case to prevent the exploitation of people or planets in the supply chains that we serve we find it relatively straightforward to recruit good technology talent, partly because we’re using a combination of technologies. Some are more mature than others but, you know, industry 4.0 Technologies clearly do attract people to try and you know for example become blockchain developers or work in machine learning. But when your business is doing something meaningful, it’s much much easier to attract talent to come and work for you because what they will do every day will make a real difference in the world.

Can you tell us what the key barriers to modernising the supply chain are? 

So the key barriers in my view to modernising the supply chain are actually the status quo. We’re used to, you know, cascading questionnaires through the supply chain for example, around responsible sourcing of conflict minerals compliance, and just returning paperwork. The entire supply chain industry is driven by bits of paper and questionnaires and Excel spreadsheets, which is the most predominant technology in the supply chain today. And although things like control tower solutions have been around for a while, and a variety of supply chain digitization tools such as ours are now coming to the fore, the status quo, what is done today is still the primary way of running supply chains. I think, overcoming that status quo and having people realise that new technologies will not only help them save money but also materially improve efficiency, as well as manage risks better will take a little bit of time.

How to build an organisation with a successful supply chain?

My view is that the role of procurement is changing. Procurement is used to be the function of buying. Increasingly procurement is now responsible for managing the supply chain, including the potential disruptions and risks within that supply chain, and driving sustainability, particularly of large manufacturing organisations. If you take for example an electric vehicle 70% of the entire carbon footprint, the manufacturer of that electric vehicle is from all of the suppliers. How can a car manufacturer achieve net-zero, if they can’t find a way to measure, and then manage down that embedded carbon within their supply chain?!

How has the COVID-19 affected the supply-chain development process?

When the world went into lockdown in March 2020 global carbon emissions fell by about 7% which is what the Paris Climate Agreement requires every year.

And why was that? 

Well, obviously we all stop travelling. But obviously, we’re at home and we’re still consuming energy and, you know, ordering things from Amazon and so global supply chains we’re still doing everything they did previously. And what it’s done is shone a light on the contribution of the global supply chain across all industries to pollution. And I don’t think we will have that brought into such sharp relief, had we not had COVID on even consumers on, you know on the street, realise that things like face masks and alcohol gel relied on all sorts of complex supply chains, and it took a while for them to gear up to meet global demand. So, it’s thrown the discipline of supply chain management into sharp relief in a way that didn’t really exist beforehand.

What will the future supply chain look like?

One word digitised.

I think digitization that has existed between for example an OEM and a tier-one for some time is going to flow down through the supply chain so that actually we will collectively have a far clearer idea of what is going on within supply chains. Procurement professionals will have the ability to manage things like the sustainability of their supply chains as part of their organization’s own NetZero ambitions.

Google released the second Android 12 Beta earlier today, and the press release emphasizes the addition of new privacy features including a privacy dashboard as well as toggles and indicators for the microphone and camera. Among other changes, there’s also a revamped Wi-Fi experience.

• Privacy dashboard: This page shows what data has been accessed on your phone and how often that data has been accessed. It has a simple timeline view of all accesses to the device’s microphone, camera, and location within the past 24 hours. This page also helps users see why an app has accessed sensitive data and make adjustments to permissions more quickly. Developers can provide an explanation for why their app handles sensitive data by responding to the new system intent ACTION_VIEW_PERMISSION_USAGE_FOR_PERIOD. Developers are also advised to track accesses in their code and in third-party libraries by taking advantage of Android’s Data Auditing APIs.
• Microphone and camera toggles and indicators: You’ll see indicators in the status bar when the microphone or camera is being used on your device. You’ll also have access to Quick Settings tiles that disable camera or microphone access. Developers are advised to review their app’s uses of the microphone or camera and remove any unnecessary uses.
• Clipboard read notification: Android 12 Beta 2 will now show a toast message at the bottom whenever an app reads the clipboard using the getPrimaryClip() method. The toast message won’t be shown if the clipboard was copied from the same app, however. To avoid being shamed on social media, developers are advised to minimize their app’s reads from the clipboard.
• Updated Wi-Fi UX experience: The second Android 12 beta has added a “simpler and more intuitive connectivity experience across the Status Bar, Quick Settings, and Settings.” There’s a new Internet panel that pops up when you long-press on the Internet Quick Setting tile (this tile was previously named “Wi-Fi” in earlier releases). This panel helps you quickly switch between networks and troubleshoot issues without diving into settings.

With the release of Android 12 Beta 2, there are now two more Betas left until the stable release later this year. The Platform Stability release will roll out in August 2021 alongside what’s presumably Beta 4. When that happens, app-facing system behaviors, SDK/NDK APIs, and non-SDK lists will be finalized.

Android 12 Beta 2 is available starting today for supported Google Pixel phones, which includes the Pixel 3, Pixel 3 XL, Pixel 3a, Pixel 3a XL, Pixel 4, Pixel 4 XL, Pixel 4a, Pixel 4a 5G, and Pixel 5. Beta 1 was available for a handful of phones from other phone makers, but Google is leaving it up to the OEM to push an update to Beta 2. The non-Pixel devices that received Beta 1 are: ASUS ZenFone 8, iQOO 7 Legend, OnePlus 9, OnePlus 9 Pro, OPPO Find X3 Pro, Realme GT, TCL 20 Pro 5G, Tecno Camon 17, Xiaomi Mi 11, Xiaomi Mi 11 Ultra, Xiaomi Mi 11 Pro, Xiaomi Mi 11X Pro / Mi 11i / Redmi K40 Pro+, and ZTE Axon 30 Ultra 5G.

Of note, Beta 2 is also available for Android TV, so developers can fire up the emulator or flash the build onto an ADT-3.

Autonomous vehicles rely on three main technologies, the first of which is high-precision localization, robust and reliable environment perception and last but not least the prediction and motion planning algorithms for safe manoeuvres. The motion planner is easier to achieve as long as the accuracy of the localization and perception data is higher.

Photo Credits: Qualcomm

• The unified localization solution

An autonomous vehicle must be able to localize itself within a few centimetres of accuracy to plan and execute complex manoeuvres required in real-world road conditions. A way to reach higher levels of accuracy is by HD Maps. Qualcomm shared:

“Our approach is to develop advanced and optimized modules that solve such challenges for our automotive OEM customers and partners.”

One such component is Qualcomm Vision Enhanced Precise Positioning (VEPP) software, which can provide lane-level positioning and a precise heading in the global coordinates by tightly fusing GNSS, IMU, camera, and vehicle odometry information in real-time and in virtually all environments.

The localization solution fuses the global pose with camera perception and HD map information and also provides an ability to generate a local map, by producing identification of gaps or errors in HD maps on the fly — all while the vehicle can operate in an autonomous mode.

• Novel techniques for efficient environment perception

One of the main focus areas for Qualcomm is to scale environmental perception technologies in a more efficient way.  Their aim is to harness all sensor types including cameras, radar, and lidar, as well as maps to build novel ways of solving environment perception challenges.

There is multisensor data fusion, which includes data association, as well as object-state estimation. Data association is basically the ability of the autonomous vehicle to correlate data from different views. State estimation involves tracking real-world objects and estimates their size, speed, and position relative to the autonomous vehicle itself.

Another Qualcomm’s focus area is the accurate prediction of the pose and tracking the size of objects on the road with 3D estimations. This happens by the use of 2D data from cameras, radar and maps. 3D estimations leveraging both deep learning and concepts from multi-view geometry. Such accurate 3D estimations can enable human-like manoeuvres for the advanced safety of the autonomous vehicle.

• Accelerating autonomous driving solutions

Qualcomm Technologies has been at the forefront of providing connectivity solutions. Building on this solid understanding of the automotive industry, the representatives of Qualcomm are now helping drive some of the biggest innovations in AD.

Their comprehensive frameworks and tools around cross-sensory calibration, active learning, big data management, and automated verification/validation can significantly reduce the time-to-market for autonomous vehicle manufacturers.