Spatial Technology Explosion: a Pandora’s Box of opportunities and risk
A quick glance at Gartner’s emerging technology hype cycle, the go to chart for any tech consultant in need of guidance, shows that more than half the trends shown rely on spatial technology or could support it.
Examples include commercial UAVs (drones), smart robots, smart dust and augmented and virtual reality to name just a few. We are no longer concerned about whether something is BIM, GIS or CAD, but rather whether or not our portable fitness device is geolocated and whether it could be tweaked to route us to avoid fitness pitfalls such as cake shops.
Spatial technologies are now not only thoroughly ubiquitous, but support highly lucrative industries. As an example, with applications in defence and law enforcement, emergency management and leisure, the VR/AR market is predicted to grow to a size of 33 billion dollars by the year 2025, serious money in anyone’s language.
Similarly, technology advances allowing for daily high resolution imagery feeds covering the entire planet are being used to support multi-billion dollar application areas. There is a heavy reliance on the ability to acquire and process massive amounts of spatial and other data. Big data can deliver big profit.
The IoT
From phones to FitBits, we now use billions of Internet of Things (IoT) devices on a daily basis. If we consider that over 80% of data generated has a spatial component, then “Big” doesn’t begin to describe the amount of spatial data IoT generates. This is even more staggering if forecasts on growth of IoT devices from 25B devices now to 100B devices by 2025 are accurate. People are rightly excited about IoT and its plethora of application areas however the IoT and wider tech-explosion also brings with it heightened privacy and security concerns for individuals and companies, from the sheer number of devices generating data to who can access the information.
Risks
Spatial and IoT
Technologies have enabled authorities to move towards Smart Cities and develop infrastructure to manage utility networks, traffic and transport systems, provide information to law enforcement agencies and support social and educational services.. However, with the proliferation in IoT devices and the location intelligence this provides and the push to drive these using AI, governance over the use of these technologies is crucial. As the FBI note, just because a device is smart it doesn’t mean it’s safe. There are several areas requiring particular consideration.
Privacy and data protection
The growing number of devices and sensors has led to a drastic increase in the amount of personal data being collected across devices. In their “Internet of Things: Privacy & Security in a Connected World” the Federal Trade Commission estimates that as few as 10,000 households can generate 150,000,000 data points per day. By willingly providing data when agreeing to terms of service when installing apps or opting in for advertising offers, people are often unwittingly agreeing to giving away potentially sensitive information. This vulnerability is easily exploited by companies to promote informed decision making. For example insurance companies could use data from fitness trackers to assist with setting premiums which while sounding reasonable becomes quite something else when applied on an individual basis.
Re-Identification of Anonymised User Data
Anonymisation refers to the stripping of personal identifiable information (PII) from a dataset such that an individual cannot be identified. It is not surprising that an industry exists that endeavors to reverse this. The ease with which re-identification of anonymized data can be carried out is directly proportional to the amount of data made available for the task and IoT collected data can certainly assist. While Yves-Alexandre de Montjoye et al, used credit card records to re-identify 90% of the individuals to whom the records pertained, the team stopped short of gaining names and addresses. They used the date, location, and price of each transaction to perform the re-identification and noted that with the addition of Facebook data (and we assume a decent geocoding service) providing names and addresses was not difficult.
Crime
From information theft to device takeover there are very real security loopholes within the IoT that leave an open door for those criminally minded. Real world cases including city-wide power grid shutdown, tampering with pacemakers and stopping vehicles on route are already surfacing. There’s even a search engine that will locate IoT devices for you. One popular feature is its ability to show you live streams from unsecured web-cams providing full HD coverage around ATM locations, building entry points and peoples bedrooms.
AI to Supervise or Not to Supervise
While IoT brings the promise of a treasure trove of big data, it is useless without big analytics. A key current state technique employed in unlocking data insights is machine learning, a subset of generalised AI. However as Elon Musk, Clive Sinclair and Stephen Hawking warn generalised AI and in particular machine learning needs to be regulated so as not to “summon the demon”. While a much larger topic in itself, examples of what could go wrong with autonomous AI include the outcome of poorly expressed goals, the inability to change these goals, and (rapidly becoming science fact) weaponisation of AI.
What should we do?
- Chris Hoar is a geo-scientist who focuses on the application of computer technology within these two fields. He is an expert in the development of IT solutions and GIS technology for urban planning, environmental and resource industry applications. Chris has worked for NGIS in Australia and Hong Kong for 18 years, project managing the implementation of spatial technology and developing and delivering solutions for many mining and oil & gas companies including Chevron Texaco, WMC and BHP. He is an active member of the GIS community, has written articles published in various journals and regularly gives presentations to the wider community on various GIS related subjects.
Related Articles
Here are more related articles you may be interested in.