Through his own attempts to program a robotic arm to play chess against him, Elliott Horowitz knows that robots have a software problem. Smart devices are getting cheaper and more widely available in different shapes and functions. Being able to tell her what to do, simply and elegantly, didn’t.
As he put it, talking with chimneyand, based on his early experience setting up a robotics lab in Viam’s New York offices, when it came to building things, “hardware people can run around the software.”
Horowitz’s new startup, Viam, is dedicated to addressing this issue and democratizing access to bots for developers. Today released a public free trial; a a program It hopes to be a one-stop shop for configuring, modeling, coding, managing and scaling bot fleets, using cloud tools that also make it easier for fleet operators to do more bot data — whether they’re analyzing a pipeline for leaks or repairing a road.
We sat down with Horowitz — who was also a co-founder of MongoDb, which is now a $12 billion market cap in the NoSQL database world — to hear more about his aspirations for the company, which has raised $30 million in Series A. Tour in February 2022, led by Tiger Global, with Union Square Ventures and Battery Ventures. (Fiam will likely rise again in 2023, he suggests, although that now doesn’t seem like a priority.)
Fiam Robot Development Kit (RDK)
“There are a lot of people who think that starting a robotics business isn’t possible,” Horowitz says. And to make them realize that it is indeed possible.”
To try to help make this possible as well as get early approval, the trial version of Viam is free to use (pay for your cloud account of course) and open source.
The “why” is simple. The Viam team would like to see robotic startups adopt Viam to help take out the friction from the software side in development, allowing them to iterate faster and bring to market robots capable of doing everything cost-efficiently from fixing potholes to catching plastic waste from the ocean.
The “how” is the product of a lot of work in progress, but it’s also deceptively simple. Viam runs on bot (through a single binary deployment under the AGPLv3 license called viam server) and in the cloud.
The platform includes three sets of APIs
- Components – eg motors, arms, GPS
- Services – computer vision, motion planning, slam
- Cloud Applications – Fleet Management & Data Management
The configuration lives in the cloud but the bots do not depend on the permanent cloud connection in any way to run. The bot code can be run directly on the bot itself or anywhere else with an internet connection.
The big idea is to “unify hardware and software prototyping” with the Viam platform that supports hundreds of popular hardware drivers and provides a software development kit (SDK) to build more.
As Horowitz told us, “We’ve defined a set of APIs for all the parts of robots that people need, like motors and arms. So there’s a gRPC API for every component you need. And you can also add your own extra tools if you need the ones we don’t.” …One of the big things we do is make it easier for hardware engineers and software engineers to work together. That means hardware engineers iterate faster,” he adds.
Once configured, the hardware can be tested in the Viam Control user interface and software engineers immediately have easy-to-use APIs to start coding. Each bot can be accessed remotely via WebRTC, and Viam natively supports high-level services such as Computer Vision, SLAM, and Motion Planning; Planning to add more in the future as well.
Elevator Step: “Centrally manage and secure each bot from anywhere. Organize bots by group and location, view each bot’s status, remotely control any bot, and maintain precise control over access to the bots—down to distinct API methods” as Viam put it in a press release for its beta launch today.
For example, users can prevent remote access to cameras deployed in sensitive environments; sync data from each bot to the cloud, integrate and analyze data in the cloud, and then publish changes to each bot. Viam can collect images, sensor readings, and even hardware metrics Through challenging network conditions and limited bandwidth. Once uploaded, users can explore and export data to run predictive analytics and train machine learning models. Trained models — or any other data — are easily pushed to bots.”
There are no points of failure in the cloud connection
For those whose Roomba machines and other small robots crashed when AWS’s US-EAST-1 data center crashed in late 2020 and then again in late 2021, cloud centralization may seem like one worrisome point of failure.
Horowitz says the platform was built to avoid this kind of issue: “We use WebRTC, which is kind of the same technology as video conferencing, and it’s all peer-to-peer. You have to have some signaling service. But that’s very powerful and doesn’t depend on No data centers because it’s just an agreement service.”
He repeats, as we make the point: “If you’re connected to the cloud, and the cloud, for whatever reason, goes down your internet connection, and a massive AWS and GCP failure, like just a complete mess, the bot doesn’t care. The bot won’t be able to add configuration updates to the cloud or Log in to the cloud or push data to the cloud. But it doesn’t care. You can turn it on, you can connect to it, you can write code against it. [locally]. “
Despite all the skepticism of those watching Elon Musk’s “Optimum” show on stage this month at a Tesla event, robots – unless you double your efforts in trying to build something human – are simply not the main obstacle to innovation as you think, he repeats: “We have a partner who is building a pipeline-checking robot for less than $1,000. The bill of materials on a robotic vacuum cleaner is $60. We’re talking to a new robotics arm company that is trying to come up with an arm that is about twice as cheap as other things like it. They don’t have to write a lot of Software because they use it. So we think we can create more competition in this space. Let hardware engineers go faster and start incentivizing some of these things. [along]! “
#Elliott #Horowitz #start #robotics #revolution #FIAM