RBR Examines Tesla's Self-Driving Beta, ARM Institute's Sewing and Textile Programs

Listen to this article

‘RBR Says’ for October 29th, 2020

Robotics Business Review reviews recent announcements, articles, research and more. Up today is Tesla’s beta program for “Feature Complete Full Self-Driving” and the ARM Institute’s robotic sewing and textile handling programs.

AUTONOMOUS DRIVING

Tesla Announces Beta Program for Feature Complete Full Self-Driving

Tesla Logo Forbes ran an excellent piece that examines Tesla’s limited beta program for “Feature Complete Full Self-Driving”. Forbes’ Senior Contributor Brad Templeton provides a review of the evolution of Tesla technology (now suitable for urban streets), and compares Tesla’s program versus competitive initiatives by Waymo, Nuro, Cruise and Zoox (Amazon). A discussion of ‘disengagement’ (changing from autonomous to driver control) follows.

Tesla’s “Thin Maps / No LIDAR” approach to autonomous driving is touched upon briefly, if only for the sake of completeness. The title of the article – Tesla’s ‘Full Self-Driving’ Is 99.9% There, Just 1,000 Times Further To Go – is a bit breathless. But the title, and subsequent coverage in the body of the piece, drives home the point that for fully autonomous driving getting to 99.9% is, for all practical purposes, nowhere at all.

Editor’s Note: For the business / technical layperson seeking a single source update on the state Tesla’s autonomous driving program, along with the technologies and issues impacting self-driving programs in general, Templeton’s Forbes piece is recommended. For those seeking additional detail and more depth, you could do worse than to seek out Templeton’s other coverage of autonomous driving.

In a nutshell, Tesla is outsourcing the testing of its imprecisely named Full Self Driving feature set to Tesla owners, and having them pay for the privilege for doing so.

RBR Says… Tesla Outsources Full Self Driving Testing to Tesla Owners
The highlight of the piece concerns Tesla’s Full Self Driving Beta Test program itself. In a nutshell, Tesla is outsourcing the testing of its imprecisely named Full Self Driving feature set to Tesla owners, and having them pay for the privilege for doing so. According to Templeton, “(Tesla) also announced a $2,000 increase in the price of the ‘FSD in the future’ package it sells today to car owners, giving them access to this software when it’s ready”.

For Tesla this approach has many advantages, starting with additional incremental revenue, a huge marketing push from Elon Musk’s Twitter feed, and scores of YouTube videos highlighting the Full Self Driving solution in action. But more importantly, the data generated by Tesla’s testers is itself of value. Tesla’s camera-based sensing solution makes heavy use of neural networks, which themselves require data at scale to be effective. With respect to data, formal (rigorous) testing and simulation only gets you so far so fast. Tesla’s Full Self Driving Beta Test program generates large amounts of additional, real-world data that can be used to improve sensing accuracy and self-driving performance, all at very little cost and possibly at a profit.

MANUFACTURING

ARM Institute Announces Eight New Robotics Technology Projects

ARM Institute The ARM (Advanced Robotics for Manufacturing) Institute, a US national consortium of 270+ member organizations focused on advancing the state of U.S. manufacturing and empower worker, recently announced the selection of eight new robotics technology projects. A total of $7.5M will be contributed across these eight projects. ARM plans to award $2.9M in project funding, and the participating organizations plan to contribute $4.6M in cost share. With the announcement of eight new robotics technology projects, all involving collaboration between government, industry, and academia, the ARM Institute’s total for completed or ‘in process’ advanced robotics for manufacturing programs increases to 59.

…textile production and apparel / garment manufacturing, which is still very much ‘hands on’ due to the difficulty of handling flexible materials… has been outsourced to a large degree to other countries where workforce costs are lower than that in the U.S.

RBR Says… ARM Rightly Emphasizing Robotics Sewing, Textile Handling and Apparel Manufacturing
The focus areas for the 59 ARM programs range widely, with most, not unexpectedly, related directly to manufacturing (logistics and workforce development also). What is notable about this announcement is that of the eight newly announced Arm Institute Robotics Technology Projects, three focus on robotics sewing and textile handling. These include:

“Bot Couture”: Robotic Clothing Manufacturing
Automated Bottom Hemming Through Robotic Garment Manipulation
Robotic Assistant for Repurposable Fabric Fusing Operations

These three programs join 3 other sewing and textile centric projects that are already underway:

Robotic Assembly of Garments (Launched 2018, HERE)
Handling and Direct 3D Draping of Limp Materials (Launched 2019, HERE)
Development of a Cost-Effective Robotic Sewing System (Launched 2019, HERE)

In total, these sewing and textile programs account for approximately 10% of ARM projects, a high percentage given that robotics automation for sewing and textile handling applications has lagged other types of robotically automated manufacturing processes.

Clearly, ARM is placing a great amount of emphasis on sewing and textile handling. The reason is completely in keeping with ARM’s charter to “accelerate transformative robotic technologies to increase U.S. global manufacturing competitiveness.” That includes textile production and apparel / garment manufacturing, which is still very much ‘hands on’ due to the difficulty of handling flexible materials, and has been outsourced to a large degree to other countries where workforce costs are lower than that in the U.S.

On a similar note, ARM is also stressing projects focused on composite manufacturing, another class of manufacturing that involves manipulation of flexible materials, and one where robotics automation, to date, has been underutilized. Example projects include:

Robot Assistant for Composites Manufacturing (Launched 2017, Completed, HERE)
Robotic Assistants for Composite Layup (Launched 2017, Completed, HERE)
Human Robot Teaming for Composite Ply Lay Up and Conforming (Launched 2018, HERE)
Collaborative Composite Sheet Layups for Complex Geometry of Small Plies (Launched 2019, HERE)

Dan Kara Vice President, Robotics, WTWH Media

Dan Kara is Vice President, Robotics at WTWH Media where he chartered with driving the company’s robotics initiatives including online and print publications, and in-person and digital events. Prior to joining WTWH, he was Practice Director, Robotics and Intelligent Systems at ABI Research and Chief Research Officer for Myria RAS, both research and advisory services firms focused on automation, robotics and intelligent systems. Dan also served as President of Robotics Trends, an integrated media and research firm serving the personal, service and industrial robotics markets. He holds an MS in Computer Science from Boston University. (Contact: dkara[AT]wtwhmedia.com, https://www.linkedin.com/in/dankara/)