From robot taxis and advanced 3D structural modelling to needle-free vaccination and safer artificial intelligence, here’s what six remarkable engineers from around the world had to say about the future tech trends to look out for.
1. Your ride’s nearly here
Robo-taxis. Cheaper sensors. Ethics. nuTonomy’s Karl Iagnemma talks about the future of cars.
The next decade, according to Ford Motor Company’s CEO Mark Fields, “will be defined by the automation of the automobile”.
The same month as Fields said this – committing the iconic American car maker to a fully-autonomous vehicle by 2021 – a world-first trial began in Singapore.
Software start-up nuTonomy began an app-based service (with engineers accompanying passengers on the free trips) in Singapore’s tiny one north business park. It aims for a driverless fleet sooner than you might expect.
“We will be conducting expanded pilots that take us much closer to our goal of launching a wide-scale ‘robo-taxi’ service in 2018,” said co-founder and CEO Dr Karl Iagnemma.
Iagnemma, also director of the Robotic Mobility Group at Massachusetts Institute of Technology, has been researching autonomous vehicles for two decades.
He believes that we “will see progress in the area of vision-based autonomous navigation” over the year. The company’s vehicles use six lidars and two dashboard cameras, as well as radar.
Lidar cost has been an area of concern for those wanting to take self-driving vehicles forward. In 2012, Google disclosed that the Velodyne system on its driverless cars cost US$70,000.
For Iagnemma, 2017 will involve expanding the Singapore trial, and dealing with each unknown as it appears.
“I view bug-finding as a sign of progress, since once a bug is detected it can generally be fixed,” he said.
“It’s the unknown bugs that keep me up at night. With that said, we have encountered many strange scenarios while driving in Singapore – a man dressed in a chicken suit recently gave our detection algorithm a bit of difficulty.”
Vehicle autonomy has long been pitched as beneficial due to the removal of human error, but there is also the topic of ethical decisions, yet to be tackled by companies like nuTonomy.
2. A future modelled by BIM
The future of construction will involve boundaries between professions dissolving.
Via six years spent literally helping redefine London’s skyline, structural engineer Roma Agrawal got a good look at the future of construction.
It’s probably no surprise that the industry is being driven by a need to build things quicker and safer. Though – as with the 60 m spire that Agrawal helped design at the top of 306 m The Shard – the trend of construction taking its cues from factories might come as a surprise.
The spire was designed in sections, built offsite, and light enough to transport and crane to the building’s top, where they were assembled. Modularity boosted efficiency, speed and safety.
“I can see this trend continuing: we’re seeing apartment buildings and student dorms being built from shipping containers or factory-made units, and there will be more of that to come,” Agrawal said.
“We’re seeing a big push on sustainability. How to reduce the carbon emissions associated with concrete? We’re recycling huge amounts of steel and using more timber.”
Agrawal is also trying to correct what she calls engineering’s ‘image problem’.
“I love what I do and it upsets me when young girls, and even boys, think engineering is messy, uninteresting and irrelevant to their lives when nothing could be further from the truth!” she said.
“My aim is to highlight the creativity that engineering entails, and make people think more deeply about how the human-made world around them came to be.”
3. Proceed with caution
Progress in artificial intelligence (AI) has been impressive in recent years, but there are concerns about what it’ll eventually be capable of.
Computers have beaten the best humans in chess (1997), Jeopardy (2011), and GO in 2016.
According to Dr Roman Yampolskiy, Associate Professor of Computer Engineering and Computer Science at University of Louisville, the latter was a standout achievement for the year in AI, coming a decade ahead of predictions.
For all the progress, there is a huge neglect of safety considerations, believes Yampolskiy, who also heads the Speed School of Engineering’s Cyber-Security Lab.
While there are thousands of papers written on safety engineering in the field, it was only this year that one was written on how to make a malevolent AI (by Yampolskiy and Federico Pistono).
Compare this to the cybersecurity world, where there’s a healthy, well-established ecosystem balance, established by hackers and security experts.
There’s also the problem that in the burgeoning field of AI, there are about 10,000 experts worldwide, though only 100 safety specialists.
“Current AI safety mechanisms are about 20 years behind current AI systems and the gap is only getting bigger,” Yampolskiy said.
Part of what concerns him about the future is the development of an artificial general intelligence (AGI), as opposed to regular AI, good at a narrow set of tasks. He is not alone, and figures including Bill Gates, Elon Musk and Stephen Hawking have shared fears AGI could jeopardise humanity.
According to some, the danger of a superintelligent computer is its lack of human goals, and the single-minded pursuit of a task would be a danger to any person in the way.
Within his field, Yampolskiy expects coming years to bring great progress in deep neural networks and “in particular applications to self-driving cars.”
Meanwhile, he will continue – as he was years before superintelligence became a hot topic – arguing the case for caution as our machines get smarter and smarter.
4. Pedal to the metal
These are pivotal years in the fast-growing use of metal 3D printing.
For many reasons, there’s a lot of excitement in the world of metal-based additive manufacturing (AM). Growth in metal AM has been recent and sharp. According to the annual industry bible, the Wohlers Report, the number of machines (by units sold) increased 54.7 per cent in 2014.
A pioneering industrial application, GE’s production of its LEAP engines (each created with 19 3D-printed nozzles) is ramping up significantly. They plan to be producing over 40,000 such nozzles annually by 2020. Also within the world of aerospace, Alcoa this year started producing titanium 3D-printed fuselage and engine pylon parts for Airbus.
Alex Kingsbury, Innovation Centre Director at the CSIRO, calls 2012 a tipping point for metal AM, but now things are really stepping into high gear. Slow, expensive and tricky to get right, the various methods of metal AM could do with serious resources to bring them closer to mainstream use.
The CSIRO scientist is hoping that the year ahead sees improvements in the size and speed of machines, both important to more widespread adoption.
It’s been said that industrial use has been especially slow to catch on in Australia, though there have been world-first medical implant applications here, notably in collaborations involving Kingsbury’s Lab 22 and Melbourne’s Anatomics.
Personalised healthcare is an area of particular promise. With implants, for example, process costs increase with the critical nature of the end part, biocompatibility of titanium alloys, and complex geometries achievable.
And she is looking forward to the proposed Innovative Manufacturing Cooperative Research Centre, which has been in limbo since 2015.
“Everyone in the industry is keen to see it get moving,” Kingsbury said.
“But I think this will be the year we see all that stuff really start to take off.”
5. Charting a course for driverless vehicles
Much of the coverage related to autonomous vehicles concerns technology or societal implications, but what about another important topic: the role of governments?
Disruptive is a term thrown around liberally nowadays. One domain where it is inarguably appropriate is to describe an autonomous car future. Exactly when self-driving cars will arrive on our roads, how quickly they become commonplace, and how they will change our lives are all unclear.
However, there are both positive and negative implications, and there’s an important role for governments to play.
This is the opinion of Lauren Isaac, Manager of Transportation Sustainability at engineering and professional services firm WSP | Parsons Brinckerhoff, and author of the recent paper Driving Towards Driverless.
Her focus is on what can be done at the local and state levels to make sure the positives are accentuated.
“People might be less willing to take public transit because if they own a driverless vehicle, they might be happy to just do a lot of trips in the driverless vehicles that they perhaps wouldn’t have otherwise taken because they’re no longer driving,” she said.
In a do-nothing scenario, a lack of proper regulation could see urban sprawl worsen, zero-occupant (errand-type) trips become normal, and vehicle kilometres and greenhouse gas emissions increase greatly.
“People might be willing to live farther from where they work because they can be in their vehicle and playing with their kids, or doing work or watching TV,” explained Isaac.
The dialogue needs to switch from the private sector leading government to the other way around.
6. No more needles
A needle-free vaccination delivery platform is challenging the 160-year-old hypodermic needle.
A mechanical engineer by training, Professor Mark Kendall moved to Oxford post-PhD, and was instrumental in developing a novel ‘gene gun’ vaccine delivery system. Since 2004 (and moving back to Australia in 2006), he has been researching, developing and commercialising the Nanopatch, a potentially disruptive needle-free vaccination delivery platform, challenging the 160-year-old hypodermic needle.
The company commercialising the Nanopatch, Vaxxas – which Kendall founded in 2011 – is an example of what Australia needs more of. The Nanopatch’s technological promise is for needle-free delivery of dry vaccine, requiring a tiny fraction compared to the syringe method. This has seen Kendall receive honours including Technology Pioneer at the World Economic Forum in 2014 and a coveted Rolex Laureate Award in 2012.
The NanoPatch, through Vaxxas, also attracted $25 million in its series B venture round last year. It uses thousands of micro-projections on a 1 cm square piece of silicon, placed to the outer layer of the skin, where populations of immune cells are much more abundant than in muscle. This means an immune response to much smaller doses compared with a needle.
Especially for developing countries, it also has the potential to do away with the expensive, and often unreliable, cold chain required to transport and store vaccines.
He said it’s hard to predict the future two decades from now but the fundamentals won’t change. The ability to probe fresh areas with confidence, ask the right questions, and apply the right principles “to learn what you need to learn” will always matter.
“It’s a credit to the engineering degrees that Australia has been producing,” Kendall said.
“If we continue that kind of thinking and commitment, I believe we should be well-placed to tackle whatever comes our way 20 years from now.”
