CYBER DUMP 54 / GHOST IN THE SHELL VR, TOYOTA HUMANOID, WAYMO TEST DRIVE
--

Welcome to Cyber Dump number 54, your look at what's happening in this insane age of technology that we live in. As always, all source links mentioned are below.

- Youtube link
- Archive.org mirror
- Torrent

VIRTUAL REALITY

This week Haptx, formerly known at AxonVR, launched their haptic feedback glove. It boasts vibration, as well as force feedback, enabled by a mini exoskeleton. Apparently this can trick the brain into feeling both the size and texture of virtual objects.

https://www.youtube.com/watch?v=2C2_kbjtjRU
https://haptx.com/

A new Ghost in the Shell themed virtual reality arcade has opened in Shinjuku, Tokyo. It's kind of like a next level laser tag setup, but now the two teams of players are fully tracked in VR, whilst they physically move around the specially built play zone.

https://www.youtube.com/watch?v=3XkupySnI7A

ROBOTICS

As a sign of things to come, robotics company Knightscope recently announced their new fleet of security robots, which can autonomously patrol, and sense the world in various ways. The 4 wheeled one looks especially futuristic.

https://www.youtube.com/watch?v=KboZXWwjCUE
https://www.youtube.com/channel/UCKyHFNyJ3QV-1rjogCYHkRA
www.knightscope.com

Toyota also announced their new humanoid robot, the T-HR3. This design boasts new torque servo modules for joints, allowing the bot to move quite flexibly and smoothly. It can also be controlled remotely by an operator.

https://www.youtube.com/watch?v=uPI0-14kEZI
http://newsroom.toyota.co.jp/en/detail/19666346/

MACHINE LEARNING

Researchers at Stanford have developed a machine-learning algorithm that can diagnose pneumonia from chest x-rays at a more consistent rate than human radiologists. It did this by analyzing a dataset of over 100,000 annotated chest x-rays, and became proficient in about a month.

https://spectrum.ieee.org/the-human-os/biomedical/diagnostics/stanford-algorithm-can-diagnose-pneumonia-better-than-radiologists

Some other researchers at UC Berkeley have also been teaching BRETT the robot how perform tasks. They used virtual reality, to physically show him how to carry out 10 tasks, slightly changing the parameters each time, and after about 30 minutes of practice, it could do them by itself.

https://www.youtube.com/watch?v=7Er84QulUBs

AUTONOMOUS VEHICLES

This month, Waymo, Google's self driving car project began testing completely autonomous car rides on public roads. One interesting point is that one of the ways they prepared for this was using neural networks to simulate 10 million miles of driving every day.

https://www.youtube.com/watch?v=aaOB-ErYq6Y
https://medium.com/waymo/with-waymo-in-the-drivers-seat-fully-self-driving-vehicles-can-transform-the-way-we-get-around-75e9622e829a

VIDEOS

The first video this week is an interview with Alex Williams, the recent winner of the 2017 Hackaday Prize. Here he talks about his underwater robot, that is capable of travelling long distances on very low power. Check out his creative buoncy mechanism.

https://www.youtube.com/watch?v=5HUCAF6898w

The next videos are a bunch of different talks, recently uploaded from this years Machine Learning conference. Many different topics covered from a range of speakers.

https://www.youtube.com/channel/UCjeM1xxYb_37bZfyparLS3Q/videos

THIS WEEK ON NODE

This week I uploaded a new video talking about the subjects of Net Neutrality, along with the importance of an open web, and what we could possible do.

https://www.youtube.com/watch?v=PriUY-zObjQ

I also made a proposal video talking about a long range, off-grid communications device that I think would be interesting to develop further. Some of you created a Slack channel to discuss this, so if one of you could post the link in the comments, that'd be great.

https://www.youtube.com/watch?v=zXTEWFb_6w4

Alright, that's it for this week. I just wanted to mention that the digital multitools are back in stock in the shop, so check them out if you want. Other than that, thanks for watching, and I'll see you in the next video.

--
BY NODE