The ethics of automation and robots has witnessed a great deal of press coverage in recent years, which backs related research, but might also end up undermining it. Current discussions in industry and policy are also endorsed by public relations and image, where the label “ethical” isn’t anything more than the new green probably used for ethics washing.
Some technologies such as cars, nuclear power, or plastics have caused political and ethical discussion and important policy efforts to monitor the course of these technologies, typically once some damage is caused. Additionally, new technologies challenge the conceptual systems and current norms.
Finally, once a new technology has been fully understood in its context, we need to shape our laws and regulations and societal response. These features are all present in the case of robotics and AI technologies, along with the more significant fear that they might end the age of human control on Earth.
For an issue to qualify as an issue for AI ethics requires not knowing what the right thing to do is. In this regard, theft, job loss, or killing with artificial intelligence isn’t a problem in ethics, but whether they are permitted under particular circumstances qualifies as a problem.
Now, let’s look at the role of automation ethics and the top challenges and their solutions in the robotics industry.
Why Ethics Matters in Automation
Digital technologies are fast revolutionizing the nature of human labor, inspiring a few workers but leaving the others behind. As a more significant number of companies embrace AI and robotics, a few jobs will be created or improved. Still, several of them are most likely to go away.
Amidst all this comes an essential question: What is the ethical approach to all this? What kind of obligations do companies have towards displaced workers in situations like these? Are there ethical ways for company leaders to attend to their workforce via digital disruption?
Researchers have struggled with those questions at MIT Technology Review’s EmTech Next Conference. They concluded that organizational leaders must better comprehend the adverse effects of the technologies they implement and commit to creating systems that drive social cohesion and economic growth.
The research paper by Pramod Khargonekar makes the case that human workers should and will always remain central and critical to the workplace of the near future, augmenting, complementing, and controlling the strengths of technological solutions.
In this regard, AI, automation, and other related technologies are merely tools that must be utilized to make human lives and our livelihoods better. Even though all this sounds aspirational, the real question is how we can get there. In their research, Sampath and Khargonekar explained four levels of what is known as the pyramid of progress:
Level 0 – Cost-focused automation
At the pyramid’s lowest level, technology is solely used to gain economic benefits by minimizing human labor. These cost-based programs aren’t only human-centric or socially conscious. The researchers stated they’re usually unable to deliver and can sometimes be unfavorable to business interests.
Level 1 – Performance-driven automation
Systems and processes are reengineered to use automation to their advantage while still using human expertise and competencies to fill in all the technological gaps. For example, in Amazon’s warehouses, workers carry out tasks that call for flexibility and agility, such as picking and packing goods.
In contrast, the robots take up routine and monotonous heavy-lifting tasks such as transporting loaded bins and totes. Even though these systems move far beyond the cost efficiencies, they’re usually driven by business metrics that do not consider the more significant implication of the staff, nor do they think about the benefits and societal costs of automation.
Level 2 - Employee-centered automation
At this particular level, the business objective is to optimize performance and enrich and develop workers. In these systems, the utmost purpose of automation is not sidelining or replacing people with machines. Instead, it is to encourage new forms of human-machine interaction that further enhance human capabilities.
On the manufacturing lines of Toyota, employees produce the goods manually at first, simplifying and innovating processes as they move forward, with machines and robots taking over only once the process is perfected.
While these approaches empower the workforce, choices and strategies are still seen from the organization’s point of view rather than in an expansive business-society ecosystem.
Level 3 - Socially responsible automation
At the pyramid’s top, automation is deployed to create a vast array of jobs for humans, driving economic success and growth while also endorsing the well-being of society. Attaining such a massive objective calls for direct and active interactions.
This means that the business leaders should devote themselves to proactively looking for new streams of revenue and job-enabling growth as they roll out and refine automation. Small business manufacturing is a segment where employees are significantly affected by the introduction of automation and robotics.
A prime example of this is Baltimore-based Marline Steel, which successfully copes with this trend. Faced with rising competition and falling demand for its products, Marlin Steel invested in automation and robotics, reengineered its production process, improved its product line with fully engineered tailored products, and extended its client base.
It did all this by equipping its workers with the expertise and training required to work in its new, cutting-edge, technology-driven place.
Giving Employees a Say in Their Layoffs
At present, the industry predominantly focuses on the lowest level, cost-savings, or probably the performance-focused level. Seeing it from that point of view, most companies are still at the very bottom row of the pyramid. This means that automation is likely to result in layoffs. Still, there is a right and a wrong way to deal with the process.
Case Example
The Finnish telecommunications giant, Nokia Corp., went through a tough time steering the marketplace for consumer electronics in the past few years. In 2008, Nokia decided to close one of its mobile production plans in Bochum, Germany, regardless of having posted a record-breaking profit for the year.
The gates of the factory were sealed shut. Employees who were just arriving for the day were sent to a local arena where they were informed that their plant wasn’t as cost-effecting. Many of their jobs were being shifted overseas. A subsequent round of boycott campaigns and public protests cost Nokia Corp. a great deal of money.
This led to a significant loss of customers, and the brand’s reputation was tarnished worldwide. This is precisely why Nokia took a different toll in 2011 when immense competition from Android and the iPhone coerced it to modify and restructure its mobile phone division.
The final decision was to shut down 8 research and development (R&D) centers and 2 assembly plants while downsizing 5 other manufacturing and assembly plants across the world. More than 1,800 workers across 13 countries were scheduled to lose their jobs.
The company then took a pledge to do the following:
- Play an active part in notifying the employees months before closure, supporting workers, and involving the affected workers in the operation and design of support programs.
- Accept its duty as a driver of local economies.
- Openly communicate with all the stakeholders, including unions, workers, local entities, and governments, even when the company didn’t have all the answers.
Workers could find another job at Nokia, apply for grants to follow a different career path, or take advantage of training opportunities to learn new skills. Career fairs were set up for the workers, and even their competitors in the local markets were invited.
Nokia even set up an incubator program that aided employees in leveraging intellectual property it no longer required into new business ventures. This move helped individual workers and led to innovation and growth in the local economies it was leaving.
Throughout this entire process, Nokia kept its employee engagement scores and factory scorecards, which usually fall during restructuring, and continued finishing R&D projects, ensuring the company’s future health.
Most relevant to automation and ethics in manufacturing is the companies implementing a stewardship mindset. They acknowledge that they are contributors to their communities, are employers, and take an active part in the lives of their customers.
10 Biggest Challenges in Robotics and Automation
The robotics industry faces numerous challenges, and the list certainly is not exhaustive. A survey regarding the challenges in robotics and automation was conducted by the journal Science Robotics . A panel of experts gradually sifted through candidates’ responses. It came up with the 10 most significant challenges that may have huge breakthroughs.
These challenges encompass enabling technologies like AI, power sources, perception, etc. Of course, ethics was also part of the list. Here are the ten biggest challenges the robotics and automation industry faces and what measures are being taken to solve them.
New materials and production methods
Motors, gears, and actuators are central to present-day robots. However, a lot of work is already being done and dusted with soft robotics, artificial muscles, and assembly strategies that’ll help create the upcoming generation of autonomous robots that are power-efficient and versatile.
Producing bio-inspired robots
Robots that are inspired by nature are becoming more prevalent in robotics labs. The primary idea is to develop robots that perform more like productive and efficient systems.
However, the most significant challenges involved with this area have lingered unchanged for 30 years – muscle-like actuators, a battery for matching metabolic conversion, autonomy in any environment, self-healing material, reasoning and computation, and human-like perception.
Materials such as actuation, couple sensing, communication, and computation must be established and shared before the segment hits off. These technological advances can lead robots with features like weight reduction, body support, morphological computation, impact protection, and mobility.
Communication in robot swarms
Robot swarms are challenging as they need to sense the environment and each robot in the swarm. They need to interact with other robots as well, all while operating autonomously. Perception action loops are essential to developing independent robots that function in unstructured settings.
Robot swarms need their communication skill to be rooted in this feedback loop. Thus, perception-action-communication loops are critical to creating robotic swarms. Currently, there are no systematic methods for doing this across enormous groups.
Increasing performances and falling prices of sensors, communications hardware, storage devices, and processors will lead to significant robot advancements in the next few years.
More efficient power sources
Usually, robots are energy-inefficient. Enhancing their battery life is a significant issue, particularly for mobile robots and drones. Thankfully, enhanced adoption of these systems leads to new kinds of battery technologies that are long-lasting, safe, and affordable.
Indeed, a lot of work is being done to make a robot’s components more power-efficient. However, the study mentions that robots that function wirelessly in unstructured settings will ultimately extract power from other sources such as mechanical movement, vibrations, and light.
Research is being conducted to enhance the battery technology by going beyond the lithium-ion and nickel-metal hydride options that are currently available.
AI that can do reasoning
According to the study, AI is called the underpinning technology for robots. However, there's still a long way for us all to copy and surpass all the aspects of intelligence we see in humans.
The key is to blend model-based reasoning and advanced pattern recognition to create AI with common sense and reason. AI that can understand and learn challenging tasks with minimal training data is also crucial.
Directing unmapped surroundings
When it comes to robots navigating and observing their environments, there’s been significant progress. For instance, take a look at self-driving cars. Navigating and mapping techniques will continue growing, but future robots must work in poorly understood and unmapped environments.
Here are some of the improvements that should be made:
- How to beat geometric maps to have a semantic understanding of the scene.
- How to learn, associate, forget memories of scenes semantically and qualitatively.
- How to persuade about new ideas and their semantic representations and come across new classes or objects in the environment through active interactions and learning.
For navigation purposes, the biggest challenge is dealing with failures and learning, adapting, and recovering. For exploration purposes, it’s developing the innate abilities to identify and create new discoveries. From a system perspective, this calls for the physical strength to withstand harsh, unsettled environments, complex manipulation, and rough handling.
The robots must possess significant autonomy levels leading to self-reconfiguration, self-monitoring, and repair such that there’s no single point of complete failure. Instead, there should be smooth system degradation.
When possible, solutions should entail control of several heterogeneous robots; use multiple assets and adaptively coordinate interface, and share information from several sources of data of variable accuracy and reliability.
Social robots for long standing engagement
In general, humans are proficient at interpreting social behavior, whereas robots aren’t. The 3 biggest challenges of designing social robots that truly communicate with humans are learning social and moral norms, modeling the social dynamics, and creating a robotic theory of mind.
Human beings are experts at interpreting and understanding social behavior that we often underestimate the difficulty of the challenge this can present for robots.
Moreover, as we are excellent at social interaction, a handful of comprehensive quantitative studies have been conducted in this area that might better allow engineers to program robots in understanding the intricacies of human interaction.
The social robots of today have been designed for short interactions, which is not how human relationships work. As a result, social robots must develop from short engagements to long-term relationships.
Brain-computer interfaces (BCIs)
Brain-computer interfaces (BCIs) allow some machines and devices to be regulated by the mind. They could be quite beneficial in enhancing human competencies in the future. However, the real challenge is to develop his technology for wider adoption.
The equipment for sensing brain signals is burdensome and costly, and data processing can be challenging. Then, there is also a long period of learning, calibration, and training involved. However, this is an interesting area to watch.
The person who lost his arm in 2005 due to cancer, Johnny Matheny, is the first-ever person to live with a progressive mind-controlled robotic arm.
Researchers from the John Hopkins Applied Physics Laboratory delivered this robotic arm to him in 2017 at his home. Over the past 10 years, Johns Hopkins received over $120 million from the US Defense Department to help pay for the development of his robotic arm.
More autonomous medical robotics
From hospital optimization and minimally invasive surgeries to prosthetics, home assistance, and emergency response, medical robots represent a rapidly developing sector. However, it can be a significant challenge to build reliable systems with greater autonomy.
A long-term challenge is to allow one surgeon to supervise a set of robots that can carry out routine procedures’ steps independently, only calling on surgeons during patient-specific, critical steps. Perhaps, the biggest challenge of automating clinical tasks is detecting, anticipating, and responding to all probable modes of failure.
Medical device regulation of independent robots will probably need to develop in a way that balances the need for verifiably safe algorithms with compliance costs. Researchers hope that AI-enabled robots can work collaboratively with surgeons and doctors and that automation can be used as dynamic implants instead of donated organs.
For these tasks, materials play a significant role. Along with creating soft robots, which will establish a safer working environment for both robots and medical staff, challenges also include reliability, security, biocompatibility, providing power, and adaptability.
Ethics
Ethics is one of the biggest challenges in the robotics industry, and it is well aware of it. These ethical problems with automation can be broken down into 5 main topics.
- AI can erode human freedom.
- Humans will no longer take account of the failures.
- Sensitive tasks that must call for human supervision can be entirely delegated to robots.
- De-skilling and unemployment of the workforce.
- Using AI in unethical ways.
As robots integrate more into society, researchers realized the importance of security and ethics. Both engineers and institutions should account for the complicated economics of bringing robots to the workforce and ensure that machines behave ethically.
Radiologists have to study images for the same reason as pilots do to successfully land airplanes so that even if the AI cannot or gets it wrong, they still can. Engineers have also been challenged with the issue that technology can be used in the wrong ways if it somehow gets into the wrong hands.
Some examples range from scanning citizens' faces in illiberal regimes to punishing law offenders or discriminating among candidates for a job. AI must be designed and utilized to always treat humans as an end and never just as a means.
Still, it’ll eventually be the mutual responsibility of lawmakers, companies, and engineers to find ways to establish responsible boundaries around the use of robotics and AI.
Follow us on social media:
