Artificial Intelligence

Panel discussion: Autonomous systems and assisted navigation through haptic feedback

14th Feb `17, 04:18 PM in Artificial Intelligence

By invitation from NASSCOM, I recently moderated a panel discussion in Cochin. The topic was Artificial Intelligence. My…

IVK
I Vijaya Kumar Contributor
Follow

By invitation from NASSCOM, I recently moderated a panel discussion in Cochin. The topic was Artificial Intelligence. My co-panelists were Dr. Roshy John of TCS and Krispian Lawrence of Ducere. The focus was on autonomous systems and assisted navigation through haptic feedback.

I opened the session with a one line description on AI, “mechanization of movement to mechanization of thought”. Summing up the journey so far, I cited examples of two iconic figures in the Artificial Intelligence space, who defined the contours of AI in the modern world. John McCarthy and Alan Kay. McCarthy, the brilliant American computer scientist is the one who coined the term ‘Artificial Intelligence’, and developed the programming language, LISP. Alan Kay, also an American computer scientist, is widely known for his work on object-oriented programming and Small Talk.

The real world use of AI faces three prime challenges- problem of embodiment, problem of meaning and problem of intrinsic motivation. The focus of the panel was on Motion Control, Autonomous Systems and Assisted Navigation through haptic feedback. Also, our discussion addressed situational awareness in real world and decision science in creating a deeper understanding of machine assisted decision making and automation.

Areas such as challenge of culture, self-organizing machines or swarms and general intelligence were not covered by this panel.

During the hour long discussion, each panelist had 5 minutes to express their opening remarks on AI and Machine learning. I asked each of the panelists their views on the following:

  • Why they chose this domain. (Aspirational query)
  • How they built their product. (Engineering query)
  • If it helped in scaling the business or if the method mattered in business. (Business query)
  • Whether AI gets premium valuation in funding rounds. (Investor query)
  • Humans versus autonomous systems. (Philosophical query)
  • What they would do differently, if they started again or if they had any advice for new entrants. (Learning/Advice query)

Key takeaways

Lechal (Ducere) started off as a navigation aid for the visually impaired. They progressed to walking trails, fitness, and the urban setting. Lechal provides guidance through haptic feedback on the foot. So the need to look at a phone is made redundant. They manufacture this product in Hyderabad and believe that India is a hotbed for hardware startups!

Dr Rosy John, who heads Robotics and Cognitive Systems at Tata Consultancy Services in Kochi, developed an autonomous Nano car. This is a personal project to showcase Software capability in the real world. The trigger was also personal, as he was deeply disturbed by a cab driver dozing off on the wheel, on the drive back from Bangalore airport.

John was stopped by Customs Officers at the airport because of this project. Why? Well, he used sensors from a German company called ‘SICK’, which made them wonder whether it made people sick!

He was stopped at times by the police as well, during the self-driving Nano trials and was often asked whether the vehicle was taking pictures of women on the road!

John believes that machines are better drivers than humans, once trained well. His wife funded his project! Perhaps, we should now rephrase the age-old quote to: ‘Behind every successful autonomous machine inventor, there is a woman!’

In conclusion, Points to ponder over:

Will a colony or a society of machines behave any differently from a society of humans? What does collective intelligence mean in this context? How about a mixed workforce of humans and machines? Will they just think and move? Do they display emotional behavior or personas?

The road ahead lies in matching human capacity and capability in terms of sensorial inputs, cognition, and motivation. While machines can match the knowledge humans possess, they may not match the ignorance of human beings. Humans are the only species which have this quality. The famous Turing test is used for testing consciousness. In the future however, we may have to test for ignorance instead!

MORE FROM BIG DATA MADE SIMPLE