Technology is everywhere. Years ago, what many pictured as science fiction have become a reality today. This is true even in the field of robotics. At the Sri Lanka Robotics Meetup, a lot of these interesting advancements come under discussion on a regular basis. The last SL Robotics Meetup discussed robotics and automation. This time, it was all about assistive robots, at the 5th edition of the robotics meetup.
Human assisting robots
First up, was Professor Kazuo Kiguchi from Japan. His discussion primarily revolved around human assisted robotics. These type of robots would essentially assist humans in various day to day chores. Professor’s focus was on 2 types of assistive robots, power assist and perception assist robots.
As the name suggests, power assist robots are utilized in scenarios where there is physically demanding tasks or in situations where a person is physically weak. So how are these particular robots controlled? The answer, EMG signals. EMG, or Electromyography, evaluates and records electrical activity produced by skeletal muscles.
These signals are what essentially enables some of the power assist robots to function. One use case would be an upper-limb power-assist robot with tremor suppression control. This is an implementation that enables better functioning for people suffering from tremors. EMG signals are affected by these tremors. Therefore, this would allow for the functioning of a power assist robot. The upper limb power assist robot basically suppresses the vibrations of the hand.
Perception assist robots, on the other hand, aims to enhance the environment perception ability of the user. Thereby, these type of robots also has similar use case scenarios to that of power assist robots. Some of these include perception assist for lower limb, sitting, and even power assist exo-skeletons.
Professor Kiguchi went on to talk about other aspects of perception assist robots as well. One scenario is where it’s able to detect danger in certain day to day situations. This would prove vital in the case of physically weak people. Following this, professor Kiguchi concluded his session, highlighting the future direction of these human assisting robots.
Brain Machine Interfaces and Assistive Robots
Following up is Dr. Thilina Lalitharatne from the University of Moratuwa. His session was all about Brain-Machine Interfaces (BMI) for Assistive Robots. In other words, how brain signals can be used for assistive robots. His research background revolved around the aging society and the disabled population. With regards to BMI, there are quite a few ways brain signal data can be obtained. These include,
- Electrocorticography (ECoG)
- functional Magnetic Resonance Imaging (fMRI)
- Electric Field Gradient (EFG)
- Near-Infrared Spectroscopy (NIRS)
- Magnetoencephalography (MEG)
In case you’re not sure what any of the above indicators mean, don’t worry. The above methods are a different way in which brain signals are obtained to feed the robot. Dr. Lalitharatne went on to share a few case studies where BMI is utilized for assistive robots. The first was a meal assistance robot.
This particular robot is primarily for patients suffering from ALS, SPI, and upper limb amputations. He briefed on how the proposed solution was derived at. Here, 430 positive images and 1300 negative images were used to train the machine. This essentially allowed the robot to identify whether the mouth is open or close. Hence, the accuracy of its functioning.
Few other BMI based assistive robots include prosthetic limbs for a person without both his limbs or even BMI for home automation/assisted living. Before concluding his presentation, Dr. Lalitharatne shared what the future will probably look like in the world of BMI based assistive robots. He mentions that this technology would play an important role in fields like self-driving cars, or more accurately, assisted vehicle driving systems.
Once the 2 presentations ended, we moved on to the final session of the day. The panel discussions, moderated by Dr. Palitha Dassanayake, talked on a number of things. This ranged from the current status of bio-robotics in the world to limitations of using BMI. Following a few curious questions from the audience, the 5th edition of the Sri Lanka Robotics Meetup officially came to a close. Until next time.