Organized by the IEEE Robotics and Automation Society Sri Lanka Section Chapter in partnership with IEEE Sri Lanka Section, the SL Robotics Meetup took place on the 28th of September 2017 at the Virtusa Auditorium located in the heart of Orion City. For those of you who don’t know, the Robotics Meetup happens at regular intervals and showcases the latest in the advancements related to robotics and related fields. The meetup consisted of two presenters along with a panel discussion. They were:
That was Prof. Chandimal Jayawardena’s opening question. He began by explaining the current challenges in this field. Service Robots are robots that provide certain services to humans. For example, there are robots designed to provide specific services such as domestic services. If you take the entire user base of people with robots, the larger share of the market is for people who use robots as pets or as a form of entertainment. As an example, you have the PaPeRo robot used for entertainment and also for childcare.
WowWee CHiP is another example as well. You also have video conference robots and healthcare robots. These have actually increased in popularity due to the shortage of trained staff. Another example was Paro. This is a robotic seel used for therapeutic services for people suffering from dementia. There are also robots equipped with various healthcare and monitoring systems used for places such as hospitals and homes.
Prof. Chandimal went on to explain that robotics is shifting from industrial robotics to service robotics. The total number of professional services such as medical robots rose by 25% from 32,939 in 2014 to 41,060 in 2015. In 2015, about 5.4 million service robots for personal and domestic use were sold, which is 16% more than 2014. The number of service robot startups has also been increased around the world. Innovation is happening, Prof. Chandima says and it’s happening around the world. For the 41,000 units sold in 2015, between 2017 and 2019, 333,000 will be produced.
Various companies produce various robots for various purposes. It is quite heterogeneous. They interact using different methods as well. Because of that, the service robotics community is facing challenges such as software development. Even though the software development landscape has changed, with regard to robotics, the principles are more or less the same with little to no advancements.
This is largely because robotics revolves around electrical engineering fields and not software development. Successful robotic applications are usually the joint effort of teams of domain experts. Prof. Chandimal explained that because of this, most robots are only adept at one particular task. As such, the results are integrated into monolithic and hardly adaptable or reusable platforms.
With the development of service robots, the way that software is developed has to change. There is no long-term support, software reusability is less because you have to develop software from scratch after an update. Modifying designs is also not as easy to do. The cost too has increased. Requirement gathering is also difficult as there’s no easy way to gather information. Requirements can only be gathered by field testing. You have a rough idea, you develop the robot, then give it to the user and only then will you find a use for it.
Prof. Chandimal then spoke about the current trends in robotics. These are component based robotics, programming tools for non-programmers, network robotics, robot software frameworks and software engineering for robotics.
Software methods and systems for robotics has increased considerably over the past decade. Pro. Chandimal also spoke about paradigms, distributed object architecture, component-based architecture and service-oriented architecture. Using an example of Robochair which is a motorized wheelchair, he explained component based robotics. According to Prof. Chandimal, apart from being a wheelchair, the device could also measure body weight, pulse rate, blood oxygen saturation, and carry out audio and video conversations as well.
From 1997, network robotics has come a long way. The interaction between robots and the cloud happened in two main approaches. The first is where developers put certain existing parts of robotic applications into the cloud. These don’t offer robotic services, but rather make it easier for developers to push existing code to the cloud for parallelization. Prime examples of these are image and object recognition.
The second approach is to host actual robotic services on the cloud. This includes services such as navigation engines and motion planning and grasping. They are essentially SaaS systems and splits the computation. Applications of cloud robotics include autonomous mobile robots, medical robots, elder care robots, personal care robots etc.
Using an example of Google Goggles, he showcased how Image recognition is used for robotics as well. By uploading and processing an image of its owner to a cloud service, a robot can greet the person in real life.
He also used an example of collective knowledge building to show how a robot collectively learned the best to grab chess pieces. They deployed multiple robots and each one learned a bit about grasping and that was shared on the cloud and they all learned from it.
With the development of IoT, it is necessary to support the interoperability amongst a large number of distributed heterogeneous robots. Prof. Chandimal spoke about the separation of behavior execution and behavior descriptions.
He also spoke about the steps that were taken and what objectives were achieved with regard to this. In conclusion, he ended his presentation to find a solution for an existing problem. Currently, there are no ways to access distributed robotic services through standard interfaces across the internet. Prof. Chandimal’s request was for someone to identify a method of accessing these services and thus put an end to the problem.
Indika Kulathunga was the second speaker for the evening. His topic was robotics and automation. He started off by defining a robot. A robot is essentially a machine that can execute an action or a set of actions based on the commands given by a human.
Automation is a combination of robots that will work together and execute a chain of tasks. He spoke about automation, specifically home automation and vehicle automation. In addition, you also have factory automation and city automation. Factory automation has advanced so much that the machinery can connect to the cloud, check inventory and stocks, daily production rate and even make intelligent choices based on the data. So you don’t have to do anything. You no longer get a machine to do a task, rather you automate it to make the entire process easier.
Vehicles now have the ability to automatically perform tasks such as parking, reversing, braking, steering, collision detection and even autopilot. While this seems to make our lives easier, it also means that we are heavily dependent on these things. For example, when was the last time you revered your car without looking at the rear camera?
Indika’s point of reference and prime example was Tesla. He used numerous examples of Tesla to showcase how automation and intelligent thinking have made driving safer and a more enjoyable experience. We’re not done, Indika emphasized, stating that in the future, cars will take into account other vehicles and objects around them. This would result in better navigation, safe roads, and even driverless taxis.
The topic then changed to home automation such as IoT, Cloud, and seamless eco-systems. These are all ways you would control the elements of your home. He emphasized that all these devices from different vendors should communicate on one platform making home automation easier. In terms of the future of home automation, he spoke about Mesh connectivity. This is where everything is connected and you can control multiple devices at the same time. This is where technology such as MIMO comes into play.
While all this is well and good, it also paves the way for potential security threats for data and sensitive information. By 2020 there will be 7.6 billion humans and approximately 50 billion connected devices. This means that you and I will seemingly have around 6 connected devices at all times. In conclusion, Indika noted that “If we’re not careful, very soon, we will be robots for robots.
The panel discussion for the Robotics Meetup consisted of the following members.
Prof Chandimal emphasized the need to improve in open standards in relation to robotics and automation. Even though it’s available, there’s no open standard. A set of standards will have to be developed. So research related to robotic ethics will also develop. Legal and ethical rules have not been developed yet. There’s nothing called reasonable judgment when it comes to robotics. An expansion to these rules can be seen. There should be a common set of standards set up in a systematic and controlled way.
Another key point that the panelists discussed was about employment. With robots seemingly taking over what humans do, jobs may be lost. On the other hand, the panelists agreed that while jobs may be lost, new jobs would also be created. After all, it’s all about creating a better way of life and making life easier. We won’t see a loss of jobs but you will see several robotic agents in the near future.
With a few more questions that were expertly answered by the panelists, this edition of the SL Robotics Meetup concluded with a small networking session accompanied by refreshments.
Subscribe to our mailing list and get interesting stuff and updates to your email inbox.
Subscribe to our mailing list and get interesting stuff and updates to your email inbox.