This week we examine a Robotic Glove designed to help those with weaker hands to recover muscle grip. We also meet Rob the Bartender and we discover a machine that can invent new mathematics. Finally we look at an AI based Dog Training system.
Robotic Glove
BioLiberty a Scottish biotech startup has designed a glove for people that suffer from hand weakness. The AI powered robotic glove helps people recover muscle grip in their hands.
The system detects their intention to grip by using electromyography which measures the electrical activity generated by a nerve’s stimulation of the muscle. An algorithm then converts the grip intent into force to help the wearer strengthen their grip on an object.
The glove will help users with a range of daily tasks from driving to opening jars. It will help people who have reduced strength in their hands due to aging or illnesses such as Multiple Sclerosis, Stroke, Arthritis, Motor Neuron Disease or Carpel Tunnel Syndrome.
There are 2.5 million people in the UK that suffer from some level of hand weakness and with our aging population that will grow. The product is in the prototype stage at the moment however the team has commenced commercialization via the Edinburgh Business School Incubator.
Machines Inventing New Math
This story involves some math that, quite frankly, I don’t understand. I know a few of my readers will and that they will no doubt provide me with feedback on the accuracy of this story. For the majority of us we should not focus on the math. We should consider the implications of what this machine can now do.
Many of the famous problems in mathematics are conjectures. A conjecture is a conclusion or proposition which is suspected to be true due to supporting evidence, but for which no proof or disproof has yet been found. For example, Fermat’s last theorem. (BTW Fermat’s last theorem was proven in 1994 after 358 years of trying, it holds the Guinness World Record for the most unsuccessful proofs).
Proposing conjectures is difficult. It must be deep enough to provoke curiosity but not so obscure as to be impossible to have a tiny glimpse of a solution. A group of researchers from the Technion in Israel and Google in Tel Aviv have developed an automated conjecturing system, called the Ramanujan Machine (named after an Indian Mathematician).
The machine is not a universal mathematics machine, it proposes formulas or conjectures, for how to compute the value of specific numbers called universal constants. An example is Pi, it is universal as it shows up all across mathematics, and constant because it maintains the same value for every circle, no matter the size. The machine has already developed a method to calculate Catalan’s constant more efficiently than any previously human discovered formulas. Thus reducing computational time and effort.
This appears to be a new breakthrough in computing. To date AI has been developed to replicate human capability but vastly more quickly and accurately. For example the glove above performs multiple calculations that a human could perform if it they had the input data and enough time to do the calculations (it is the speed of the calculation and the ability to have it feed into the robotic part of the glove in real time that makes the product work).
The Ramanujan Machine proposes new potential solutions to a problem. This may mark a new progress point in computing, somewhat like the first time that computers beat the chess masters (which happened in 1997).
The paper outlining this new machine can be found in Nature here.
A Robotic Bartender
When MSC Cruises once again sails the seven seas, Rob the Humanoid Bartender will be serving drinks in Virtuosa’s Starship Club. No idea why they called their robot bartender Rob, but the name seems appropriate.
To order a drink, guests place their order in “vertical digital cockpits” (I think they mean that guests will use the screen on the wall or on their table) and Rob will mix a range of alcoholic and non-alcoholic drinks. Rob speaks 8 languages and each drink is served with a long story about his skiing exploits before he retired and became a bartender (I added that last bit but it seems appropriate for a bartender called Rob).
Rob has an LED face to allow him to display human emotion, particularly his sad face when nobody tips the robot bartender.
Rob is not the first robot bartender. Glacierfire in Iceland is a bar built around their robot bartender, in Spain Macco Robotics’ robot serves beer and in Las Vegas the Tipsy Robot slings drinks. Rob debuts in April 2021 when the Virtusoa restarts cruising in the Mediterranean.
AI Dog Training
Sometimes it is difficult to explain exactly what you want your dog to do without him just giving you a blank look.
Colorado State University students have developed an AI based solution to help solve this problem. Using NVIDIA’s Jetson’s edge AI platform they trained an image classification system to determine a dog’s position. The AI was trained with over 20,000 images of dogs in different positions.
The system works by giving the dog a command. If the dog responds to the command by adopting the correct posture, the machine dispenses a treat. I am sure the system works well for simple commands like, sit, lie down etc. however how do you teach the dog more complex behavior. Such as to fetch your paper or get a beer from the fridge? Every home may still need a “Rob the Bartender”.
Paying it Forward
If you have a start-up or know of a start-up that has a product ready for market please let me know. I would be happy to have a look and feature the startup in this newsletter. Also if any startups need introductions please get in touch and I will help where I can.
If you have any questions or comments please email me via my website craigcarlyon.com or comment below.
I would also appreciate it if you could forward this newsletter to anyone that you think might be interested.
Till next week.