We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon
Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
For the first time, Amazon today showed off its newest fully electric delivery drone at its first re:Mars conference in Las Vegas. Chances are, it neither looks nor flies like what you’d expect from a drone. It’s an ingenious hexagonal hybrid design, though, that has very few moving parts and uses the shroud that protects its blades as its wings when it transitions from vertical, helicopter-like flight at takeoff to its airplane-like mode.
These drones, Amazon says, will start making deliveries in the coming months, though it’s not yet clear where exactly that will happen.
What’s maybe even more important, though, is that the drone is chock-full of sensors and a suite of compute modules that run a variety of machine learning models to keep the drone safe. Today’s announcement marks the first time Amazon is publicly talking about those visual, thermal and ultrasonic sensors, which it designed in-house, and how the drone’s autonomous flight systems maneuver it to its landing spot. The focus here was on building a drone that is as safe as possible and able to be independently safe. Even when it’s not connected to a network and it encounters a new situation, it’ll be able to react appropriately and safely.
When you see it fly in airplane mode, it looks a little bit like a TIE fighter, where the core holds all the sensors and navigation technology, as well as the package. The new drone can fly up to 15 miles and carry packages that weigh up to five pounds.
This new design is quite a departure from earlier models. I got a chance to see it ahead of today’s announcement and I admit that I expected a far more conventional design — more like a refined version of the last, almost sled-like, design.
Besides the cool factor of the drone, though, which is probably a bit larger than you may expect, what Amazon is really emphasizing this week is the sensor suite and safety features it developed for the drone.
Ahead of today’s announcement, I sat down with Gur Kimchi, Amazon’s VP for its Prime Air program, to talk about the progress the company has made in recent years and what makes this new drone special.
“Our sense and avoid technology is what makes the drone independently safe,” he told me. “I say independently safe because that’s in contrast to other approaches where some of the safety features are off the aircraft. In our case, they are on the aircraft.”
Kimchi also stressed that Amazon designed virtually all of the drone’s software and hardware stack in-house. “We control the aircraft technologies from the raw materials to the hardware, to software, to the structures, to the factory to the supply chain and eventually to the delivery,” he said. “And finally the aircraft itself has controls and capabilities to react to the world that are unique.”
What’s clear is that the team tried to keep the actual flight surfaces as simple as possible. There are four traditional airplane control surfaces and six rotors. That’s it. The autopilot, which evaluates all of the sensor data and which Amazon also developed in-house, gives the drone six degrees of freedom to maneuver to its destination. The angled box at the center of the drone, which houses most of the drone’s smarts and the package it delivers, doesn’t pivot. It sits rigidly within the aircraft.
It’s unclear how loud the drone will be. Kimchi would only say that it’s well within established safety standards and that the profile of the noise also matters. He likened it to the difference between hearing a dentist’s drill and classical music. Either way, though, the drone is likely loud enough that it’s hard to miss when it approaches your backyard.
To see what’s happening around it, the new drone uses a number of sensors and machine learning models — all running independently — that constantly monitor the drone’s flight envelope (which, thanks to its unique shape and controls, is far more flexible than that of a regular drone) and environment. These include regular camera images and infrared cameras to get a view of its surroundings. There are multiple sensors on all sides of the aircraft so that it can spot things that are far away, like an oncoming aircraft, as well as objects that are close, when the drone is landing, for example.
The drone also uses various machine learning models to, for example, detect other air traffic around it and react accordingly, or to detect people in the landing zone or to see a line over it (which is a really hard problem to solve, given that lines tend to be rather hard to detect). To do this, the team uses photogrammetrical models, segmentation models and neural networks. “We probably have the state of the art algorithms in all of these domains,” Kimchi argued.
Whenever the drone detects an object or a person in the landing zone, it obviously aborts — or at least delays — the delivery attempt.
“The most important thing the aircraft can do is make the correct safe decision when it’s exposed to an event that isn’t in the planning — that it has never been programmed for,” Kimchi said.
The team also uses a technique known as Visual Simultaneous Localization and Mapping (VSLAM), which helps the drone build a map of its current environment, even when it doesn’t have any other previous information about a location or any GPS information.
“That combination of perception and algorithmic diversity is what we think makes our system uniquely safe,” said Kimchi. As the drone makes its way to the delivery location or back to the warehouse, all of the sensors and algorithms always have to be in agreement. When one fails or detects an issue, the drone will abort the mission. “Every part of the system has to agree that it’s okay to proceed,” Kimchi said.
What Kimchi stressed throughout our conversation is that Amazon’s approach goes beyond redundancy, which is a pretty obvious concept in aviation and involves having multiple instances of the same hardware on board. Kimchi argues that having a diversity of sensors that are completely independent of each other is also important. The drone only has one angle of attack sensor, for example, but it also has a number of other ways to measure the same value.
Amazon isn’t quite ready to delve into all the details of what the actual on-board hardware looks like, though. Kimchi did tell me that the system uses more than one operating system and CPU architecture, though.
It’s the integration of all of those sensors, AI smarts and the actual design of the drone that makes the whole unit work. At some point, though, things will go wrong. The drone can easily handle a rotor that stops working, which is pretty standard these days. In some circumstances, it can even handle two failed units. And unlike most other drones, it can glide if necessary, just like any other airplane. But when it needs to find a place to land, its AI smarts kick in and the drone will try to find a safe place to land, away from people and objects — and it has to do so without having any prior knowledge of its surroundings.
To get to this point, the team actually used an AI system to evaluate more than 50,000 different configurations. Just the computational fluid dynamics simulations took up 30 million hours of AWS compute time (it’s good to own a large cloud when you want to build a novel, highly optimized drone, it seems). The team also ran millions of simulations, of course, with all of the sensors, and looked at all of the possible positions and sensor ranges — and even different lenses for the cameras — to find an optimal solution. “The optimization is what is the right, diverse set of sensors and how they are configured on the aircraft,” Kimchi noted. “You always have both redundancy and diversity, both from the physical domain — sonar versus photons — and the algorithmic domain.”
The team also ran thousands of hardware-in-the-loop simulations where all the flight services are actuating and all the sensors are perceiving the simulated environment. Here, too, Kimchi wasn’t quite ready to give away the secret sauce the team uses to make that work.
And the team obviously tested the drones in the real world to validate its models. “The analytical models, the computational models are very rich and are very deep, but they are not calibrated against the real world. The real world is the ultimate random event generator,” he said.
It remains to be seen where the new drone will make its first deliveries
Amazon.com, Inc. (AMZN) | Stock Discussion ForumsShare
What Amazon Might Want With Boost Mobile wired.com Why might the retail-and-cloud-computing giant be interested in buying a wireless network? In the near term, Boost would give Amazon another service to sell. The company has been expanding its portfolio of Amazon-branded products, from batteries to apparel to kitchenware. Amazon already sells phones and, via third-party sellers, prepaid SIM cards. It's not much of a stretch to think that Amazon could sell phones bundled with an Amazon-branded wireless service.
Longer term, the acquisition could tie into Amazon’s efforts to gain more control over the networks it relies on to reach customers. Everything from its Echo devices to its Prime Video streaming service to its potential delivery drones relies on internet connectivity provided by other companies.
Amazon.com, Inc. (AMZN) | Stock Discussion ForumsShare
-- Courier says it won’t renew contract for U.S. air services
-- Focus will be on other customers such as Walmart, Target
FedEx Corp. said it wouldn’t renew its U.S. air-delivery contract with Amazon.com Inc., paring a key customer relationship as the largest online retailer deepens its foray into freight transportation.
The delivery giant will instead focus on “serving the broader e-commerce market” with U.S. package volume from online shopping expected to double by 2026, according to a FedEx statement Friday. The Amazon contract expires at the end of this month, and doesn’t cover international services or domestic operations at other units such as FedEx’s ground deliveries.
FedEx’s surprise move signals that the No. 2 U.S. courier will bank on smaller e-commerce customers that lack Amazon’s bargaining power for big volume discounts. Amazon’s emergence as a logistics giant is piling pressure on FedEx and United Parcel Service Inc. for cheaper and speedier deliveries, as the e-commerce powerhouse builds its own aircraft fleet and delivery capabilities.
“They know their Amazon business is going to continue to shrink,” said Satish Jindel, founder of SJ Consulting Group, referring to FedEx. “Why have your capacity be used up by a customer that’s going to continue to chip away? They’d rather use that capacity for other customers.”
FedEx rose less than 1% to $158.17 at 2 p.m. in New York. The shares erased gains on the company’s announcement about Amazon before recovering some of the lost ground. Amazon held steady, trading 2.7% higher at $1,801.10.
“We respect FedEx’s decision and thank them for their role serving Amazon customers over the years,” the Seattle-based retailer said in a statement.
FedEx said Amazon represented 1.3% of sales last year. Jindel estimated that FedEx’s domestic air-parcel business with Amazon is probably “a few hundred million a year, at the best.”
The Memphis, Tennessee-based courier said Amazon wasn’t its largest customer. In an email, FedEx said it would focus on customers such as Walmart Inc., Target Corp. and Walgreens Boots Alliance Inc.
“There is significant demand and opportunity for growth in e-commerce, which is expected to grow from 50 million to 100 million packages a day in the U.S. by 2026,” FedEx said in the statement. “FedEx has already built out the network and capacity to serve thousands of retailers in the e-commerce space.”
Earlier this year, Amazon announced its Scout sidewalk delivery robot. At the time, details were sparse, except for the fact that the company had started to make deliveries in a neighborhood in Washington State. Today, at Amazon’s re:Mars conference, I sat down with Sean Scott, the VP in charge of Scout, to talk about how his team built the robot, how it finds its way around and what its future looks like.
These relatively small blue robots could be roaming a sidewalk near you soon, though as of now, Amazon isn’t quite ready to talk about when and where it will expand its network from its single neighborhood to other areas.
“For the last decade, we’ve invested billions of dollars in cargo planes and delivery vans, fulfillment center robots, and last holiday period, we shipped over a billion products with Prime free shipping,” Scott told me. “So it’s my job as VP of Amazon Scout to bring another new, innovative, safe and sustainable solution to this delivery network to help us really grow quickly and efficiently to meet customer demand.”
Currently, in Amazon’s trial, the robots are always accompanied by human assistants. Those assistants — and they probably look a bit like robot dog walkers as they trot through the neighborhood — are currently the ones who are taking the packages out of the robot when they arrive at their destination and put it on the customers’ doorsteps. For now, that also means the customers don’t have to be home, though chances are they will have to be once this project rolls out to more users.
As of now, when it’s ready to make deliveries, Amazon drives a large van to the neighborhood and the Scout robots leave from there and return when they are done. Scott wouldn’t say how far the robots can travel, but it seems reasonable to assume that they could easily go for a mile or two.
As we learned earlier this year, Amazon did make a small acquisition to kickstart the program but it’s worth stressing that it now does virtually all of the work in house, including building and assembling the robots and writing the software for it.
“For Scout we’re actually owning the entire development from the industrial design to the actual hardware, mechanical, electrical, the software, the systems, manufacturing and operations,” said Scott. “That really helps us control everything we’re doing.” Having that end-to-end control enables the team to iterate significantly faster.
The team even built a rig to test the Scout’s wheels and in the process, learned that the wheels’ material was actually too soft to survive the rigors of daily sidewalk driving for long.
Inside its labs, the team also built a sidewalk environment for real-world testing and did most of the initial training in the real world but also heavily relies on working with simulations now. Indeed, since there are basically no maps for navigating sidewalks, the team has to build its own maps of every neighborhood it goes into and it then uses this highly detailed map in its simulation.
That’s important, Scott noted, because simply using a game engine with repeating textures just wouldn’t be good enough to train the algorithms that keep the robot on track. To do that, you need real-world textures, for example.
“We thought about building a synthetic world, but it turns out building a synthetic world is much harder than copying the real world,” Scott said. “So we decided to copy the real world.” He showed me a video of the simulated robot moving through the simulation, using a map that looks a bit like a highly zoomed-in Google Maps 3D view. Not perfect, but perfectly reasonable, down to the gutters on the street and the small bumps where two concrete plates on the sidewalk line up.
This simulation allows Amazon to make thousands of simulated deliveries before the team ever goes out to test the robot on the street. In the demo I saw, the robot had no issues navigating around obstacles, pausing for crossing cats and getting to his destination. That’s possible thanks to a combination of detailed maps and high-resolution imagery of its surroundings, combined with GPS data (when available) and cutting-edge machine-learning techniques.
Once it is out and about, though, the robot will have to face the elements. It’s watertight, something you’d expect from a company that is based in Seattle, and it’s got sensors all around to ensure it can both find its way on sidewalks that are often littered with obstacles (think trash day) and full of curious cats and dogs. Around the robot is an array of cameras and ultrasonic sensors, all of which are then evaluated by a set of machine learning algorithms that help it plot its path.
“We jokingly refer to the sidewalk as the Wild West,” said Scott. “Every sidewalk is a snowflake and every neighborhood is a collection of snowflakes.”
At times, the robot also has to deviate from the sidewalk, simply because it is blocked. In those cases, it will opt for driving on the street. That’s something local laws in many states now allow for, though Scott tells me that the team only considers it when it’s a street where a pedestrian would also feel comfortable. “If you feel safe walking on that road, that’s where we want to be. We want to be viewed as a pedestrian and treated as a pedestrian,” he said. And that’s how the law in Washington State looks at these robots, which, for example, mean that they have to be given the right of way.
Scott also noted that the team designed the robot so it would be visible when necessary, with blinking lights when it crosses a street, for example, but also a bit boring, so that it would blend into the environment. “We really want this to blend into the background and be part of the environment and not be this loud and obnoxious thing that’s always rolling through the neighborhood,” said Scott. So it has the bright blue Amazon Prime color on top to be seen, but is otherwise relatively bland and without any anthropomorphic features. It’s just your average neighborhood delivery robot, in other terms.
As it moves along, it makes very deliberate movements, which Scott believes will make people feel more comfortable around it. Unlikely a drone, there’s no major risk when any parts of the robot break during a mission. Somebody can simply come and pick it up. Still, the team says it did design the robot with safety at the front and center of its process.
One thing that’s currently not clear — and that Amazon didn’t want to talk about yet — is how it will solve the actual handover of the package. Right now, the assistant handles this part, but in Amazon’s photos, the customer walks up to the robot and takes the package out of it. That’s a reasonable scenario, I think. In the long run, Amazon could also outfit the robot with multiple compartments to make multiple deliveries in one go.
One advantage of the robot has over human delivery people is that if you’re not home, it can just wait for a while, Scott said. So it’s conceivable that you’ll come home one day and there’s a Scout, standing patiently in front of your door, waiting to deliver your latest impulse order. Until then, it’ll likely be a while, though. Amazon won’t commit to any timetable or wider rollout.
AWS costs every programmer should know Jun 9, 2019 • David Hatanian
The title for this blog post is a direct reference to Latency Numbers Every Programmer Should Know. There are several versions of those numbers available now, and I could not find the original author with certainty. Some people attribute the original numbers to Jeff Dean.
When working on a project that will reach a certain scale, you need to balance several concerns. What assumptions am I making and how do I confirm them? How can I get to market quickly? Will my design support the expected scale?
One of the issues associated with scale, is the cost of your infrastructure. Cloud providers allow you to provisions thousands of CPUs and store terabytes of data at the snap of a finger. But that comes at a cost, and what is negligible for a few thousand users might become a burning hole in your budget when you reach millions of users.
In this article, I’m going to list reference numbers I find useful to keep in mind when considering an architecture. Those numbers are not meant for accurate budget estimation. They’re here to help you decide if your design makes sense or is beyond what you could ever afford. So consider the orders of magnitude and relative values rather than the absolute values.
Also consider that your company may get discounts from AWS, and those can make a massive difference.
Compute What’s the cost of a CPU these days? I used the wonderful ec2instances.info interface to extract the median price of a vCPU.
With 1 year convertible reservation (all up front)
With 3 years convertible reservation (all up front)
With spot pricing (estimated)
I estimated spot pricing based on anecdotal data I got from various sources. As the prices vary within a day and I could not find a reliable data source for it.
AWS represents the computing power of its machines in Elastic Compute Units, and 4 ECUs represent more or less the power of a modern CPU. So the prices above are for one CPU or core, not one instance.
Here’s the price of 1 ECU in $ per hour across all instance types I looked at:
And here’s how on-demand compares with 1 year and 3 year reservations (both convertible, upfront payments):
Storage So you want low latency, high throughput and are planning to store everything in Redis? Then on top of those CPU costs, you’ll need to pay for RAM.
I used the same approach to extract the median price of 1GB of RAM on EC2. Elasticache is more or less twice as expensive for on-demand but prices drop quite quickly when looking at reserved instances.
Median monthly cost
1 GB RAM
1 GB RAM 1 year convertible reservation (all up front)
1 GB RAM 3 years convertible reservation (all up front)