Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

The mention of big data can all too quickly evoke chrome utopias with everything from your watch to the park bench connected to the internet and harvesting information. Yet, ultimately, the benefits of big data come from simply connecting information to the people to whom it is most valuable — and that can be done without next-generation machines.

Collecting data

Many big data projects are carried out with more mundane means, where data is collected not by devices but by hand. A public spending watchdog in Nigeria, for example, uses face-to-face meetings with Members of Parliament to collect spending data.

Through talking to MPs directly, BudgIT convinces them to share hard-to-get information about what their departments are really spending money on. Nigeria is ranked 136th out of 176 countries on Transparency International’s Corruption Perceptions Index, and MPs are suspicious of BudgIT’s requests for spending information for fear of being investigated.

The spending information can’t be collected automatically because Nigeria’s MPs are often unwilling to share their data in the first place. BudgIt’s staff, therefore, have to convince them that the data will be used to serve the public and not to prosecute them personally. This process makes the data more reliable than the reports that had previously been released. BudgIT’s efforts helped persuade the Nigerian government to join the Open Government Partnership, a multilateral initiative to promote transparency and stamp out corruption.

Releasing data

While the most common way of releasing data to the public is via online portals, one startup in Ghana sends representatives armed with information into communities to speak with people directly.

TransGov collects data on development projects such as roads and schools, and distributes it to people in the area — not only online, but also via text message, word of mouth and voice recognition technology, which allows illiterate Ghanaians to find the information they need simply by speaking into their phones. TransGov’s aim was to improve democratic participation in the country: when it formed in 2015, staff discovered that 80% of the Ghanaians they surveyed did not know who their local assembly representative was. By bringing relevant information to the localities via public meetings, TransGov staff are able to engage people with the political system using the data that matters most to them.

Since only 28% of Ghanaians have access to the internet, TransGov is unable to use its website alone to distribute their data. By sending staff out to rural areas, they access those citizens who would otherwise remain ignorant of political processes in the country. The combination of methods has helped TransGov reach 400,000 Ghanaians so far.

Low-tech solutions in the developed world

Even in some of the world’s most advanced cities, certain data collection tasks still need to be completed manually. In 2007, New York City’s Department of Health and Mental Hygiene (DOHMH) began collecting masses of data on its rat population by sending its officers out to make in-person inspections.

There are few technologies able to accurately track the presence of rats, as sensors can’t be placed in every building in the city. DOHMH has to send its officers out to check areas property by property for rat activity. For years, they only conducted rat inspections after members of the public made complaints about particular properties. In 2007, they inspected all the properties in the Bronx unprompted, in order to build a map showing where rat infestations were. They then published this data via an online portal, allowing local residents to tackle the problem themselves by cleaning up neighbourhoods and removing the rats’ food sources.

By successfully coordinating cooperation between residents and agencies, the DOHMH has helped to reduce rat activity by 30% to 60% in certain New York neighbourhoods.

Low-tech can make big data, but costs in the long term

In these examples, manual collection or distribution of data helps to navigate technological or political difficulties. Whether due to suspicious MPs, poor internet connections or the sheer difficulty of tracking rat infestations with machines, each organisation had to develop a way to collect and distribute data by hand.

The irony is that a lack of technological infrastructure — which necessitates these cheaper, low-tech solutions — is more expensive in the long run. The fact that swathes of land remain unconnected to the internet has forced companies in Africa like BudgIT and TransGov to improvise solutions to the problems in front of them. While this limits their start-up and operations costs, as time passes this manual labour becomes ever more expensive.

Collecting and distributing data by hand takes a lot of time and costs money. The savings offered through collecting data with automated systems and publishing them online almost always outweigh the cost they take to implement. In these three cases, they just need to wait for technology or politics to catch up.

This article was originally published on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

How to do big data the low-tech way was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

By 2010, the world was producing as much data every two days as it had done between the start of civilization and 2003, while 90% of all data has been created within the past two years.

This exponential growth in information is transforming what governments can do for the people they represent. Authorities are now predicting accidents before they happen, seeing where criminals will strike before they do, and consigning lengthy form-filling to the past.

At Apolitical, we have identified the five most significant ways that information sharing and predictive analytics are transforming government services around the world:

1. Bridging political divides

Taiwan is using social media to crowd-source legislation, breaking a six-year political deadlock on online alcohol sales in the process.

The system, known as vTaiwan, finds areas of public agreement and frames legislation around them. The public is invited to post principles that should underpin legislation into discussion threads distributed via Facebook. Unlike conventional forums, users rate others’ suggestions rather than commenting on them. Responses are then clustered, showing statements that cut across political divides.

After further consultation and review by experts, the government either implements a bill based on the responses or explains why it is not doing so.

2. Predicting crime before it happens

By linking buildings with types of crime, police can see where the next robbery, assault, arson or stabbing is likely to occur before it happens.

The method, known as Risk-Terrain Modelling, establishes correlations between types of crime and particular buildings. All addresses of potentially ‘risky’ buildings are logged on a digital map, which is overlaid with the locations of past crimes. Buildings are ranked according to the number of incidents that have occurred close by, revealing combinations of structures that pose the greatest risk. Areas where these combinations occur are flagged to police, letting them focus on small areas likely to attract large volumes of crime.

Where it was implemented, gun crime dropped by 35% in Newark; Colorado Springs cut vehicle theft by 33% and Glendale, Arizona, reduced robberies by more than 40%.

Developed by Professors at Rutgers University, the system is currently being used in more than 30 countries and over 30 US states.

3. No more filling out forms

In Estonia, citizens can complete their taxes in just five minutes or incorporate a business in less than 20, and are never asked to submit the same information to the government twice. This is thanks to the country’s data exchange network, X-Road, which saves more than 240 hours of work every three minutes by automatically transferring information between government agencies.

X-Road lets approved databases request and share information without human commands. The process allows data to be automatically retrieved and made available to officials as it is needed for specific tasks, without them having to make any manual request.

The technology underpinning the X-Road system is open source and Estonia is already helping one other country, Finland, adopt it.

4. Hospitals could predict A&E arrivals ten years in advance

As hospital waiting times continue to lengthen, Australia might just have the answer: earmark beds for patients before they even walk into A&E. Across Queensland, hospitals are using a tool that tells hospitals how many patients will arrive and what injures they will have, hours, days, weeks, months or years in advance.

The tool identifies subtle patterns in historical admission and discharge data, allowing hospitals to anticipate injuries long before they happen. This lets hospital staff set aside beds for patients with broken-legs, alcohol poisoning, pneumonia and other injuries when planning surgeries, preventing last minute cancellations or lengthy patient waits. Alternatively, staff can predict arrivals by hour, letting them know how many beds are required at any given moment.

In use in around 50 hospitals across Queensland, the tool is saving money and lives. The state’s hospitals are saving up to $2.5 million (USD) each year while the value to the state from improved patient outcomes could be as high as $77.5 million. Waiting times have also fallen, in one case by as much as 20%.

5. Predicting car crashes before they happen

If you prang your car in the coming years, you might find that the emergency services set off before you have even dented your fender.

A new crash prediction system, set to be trialled by Las Vegas, uses artificial intelligence to identify correlations between the time and locations of previous car accidents with factors such as weather, traffic flow, construction sites and road hazards. Developed and tested in Israel, the technology predicted accidents and jams with up to 70% accuracy, and provided warnings of likely collisions up to two hours in advance.

As well as telling officials when accidents will happen, the system may also be able to show them the best way to prevent them. This could involve altering road signals, sending out patrol vehicles or even preparing emergency services.

This article was originally published on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

Five ways government is using big data to make your life better was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

Cincinnati is showing citizens how public services and policy are affecting where they live through easily digestible, interactive maps and graphs that visualise city data. The CincyInsights portal lets users track public spending, business activity, ambulance calls, rubbish pickups and other services specific to their street and neighbourhood. The platform has significantly increased the use of and demand for public data, as well as improved public services: its heroin dashboard played a key role in helping medical officials combat the city’s opioid crisis.

Results & Impact

The CincyInsights data visualisations are making it easier for citizens to understand how policy affects public services where they live, and helping officials better use data to improve services. The platform’s Heroin tracker, for example, played a key role in helping medical staff respond to the city’s opioid problem by showing them where and when incidents were occuring, allowing them to preemptively deploy medical services to worst-affected areas. CincyInsights’ street-level data has also significantly increased interest in and use of the city’s open data, with citizens now downloading substantially more data since its launch, according to Leigh Tami, Chief Performance Officer and Director at Cincinnati’s Office of Performance and Data Analytics.

Key Parties

Cincinnati Office of Performance and Data Analytics, Tableau

How

CincyInsights uses visualisation software from Tableau to create data-driven graphs and interactive maps. Each dashboard (an interactive, at-a-glance visualisation of key data points) is connected to the city’s data warehouse. Created as part of the CincyInsights project, the warehouse pulls in data from city agencies and private vendors. This occurs at set time intervals, varying from every five minutes for the snow plough tracker to once a day for other data sets, and is often done via APIs. Before this takes place, city officials identify the type of data they want to display from each data set and clean it accordingly. Links between data sets are also planned in advance, allowing maps to integrate information from more than one source. The code is then written to extract the information in the desired format and store it in Cincinnati’s data warehouse. A second set of code pushes the data from the warehouse into maps and graphs on the CincyInsights platform. The data is automatically refreshed whenever new information is pulled into the data warehouse.

Hurdles

Creating the data infrastructure to feed the visual platform was complex and time-consuming, primarily because of the extensive cleaning and standardisation needed to make the data useable. For instance, data sets often contained personal information that needed to be anonymized, such as officer numbers in the case of police data. Different data sets sometimes also recorded the same parameters using different terms, each of which needed to be standardized. Additionally, all links between data sets had to be identified manually. For instance, to allow users to move between viewing information by addresses and neighbourhoods, code needed to be written to link those data sets. The visualisation process, therefore, requires considerable foresight to envision all the ways in which data needs to be combined, and ensure all discrepancies are ironed out.

Read the full article on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

Cincinnati makes maps to show people how government policies affect where they live was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

Nigeria has partnered with a public spending watchdog to improve government accountability and transparency. The NGO BudgIT tracks budgets and proposed spending across governments and departments, highlighting anomalies in easily disseminated graphics and asking citizens to track big projects by uploading pictures of their progress. Nigeria is ranked 136th out of 176 countries on Transparency International’s Corruption Perceptions Index.

Results & Impact

The organisation has reached 750,000 Nigerians online and offline, and gets 2,000 data requests monthly from individuals, private companies and development agencies. BudgIT played a key part in convincing the Nigerian government to join the Open Government Partnership, a multilateral initiative to promote transparency and stamp out corruption.

Key Parties

Nigeria’s National Assembly, Nigerian national and state governments, BudgIT.

How

The NGO BudgIT mines public spending data, then visualises findings in reports and infographics to simplify budgets for Nigerians. It provides critical information on issues that directly affect citizens like urban development, housing and roads, and asks citizens to track public projects by uploading pictures of their progress. The project is funded by philanthropic foundations Gates and Omidyar.

Hurdles

The biggest challenge is earning trust from members of parliament who remain suspicious of the project. MPs tend to take requests for clarification as a personal attack, but BudgIT’s ongoing work in the country has helped ease tensions: “People in government have come to realise that we’re not in the business of trying to jail people,” said co-founder Oluseun Onigbinde.

Read the full article on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

Nigeria holds public servants accountable with budget watchdog was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

Indonesia is using crowdsourced language data to implement policies directed at achieving the Sustainable Development Goals. The Translator Gator online game incentivises citizens to translate words related to the SDGs from English into their native language by rewarding users with phone credit. Indonesians have submitted 109,000 translations, which government is using to create taxonomies for six of the country’s 300 languages and monitor public opinion on SDG-related policies.

Results & Impact

In four months, the Translator Gator web game received 109,000 contributions, at least 97% of which are accurate. The language data has helped create taxonomies for six of Indonesia’s native languages and drive policy research.

Key Parties

Indonesia’s Ministry of National Development and Planning, the United Nations, Pulse Lab Jakarta

How

Indonesia’s Ministry of National Development and Planning and the United Nations partnered to create Pulse Lab Jakarta, an innovation hub that uses data analysis to promote social and economic development in Indonesia. Pulse Lab built Translator Gator, a game that prompts citizens to translate words and phrases related to the SDGs from English to their native languages.

In addition to translating words, users rank others’ interpretations, suggest synonyms and sort words and phrases into categories. Users are rewarded for their translations, which they can cash in for phone credit. The native languages used in the game include Jawa, Sunda, Minang, Bugis and Melayu.

Hurdles

The biggest challenge for Pulse Lab is encouraging a diverse cross-section of people in 11 countries to use the app. In an effort to make the game more attractive to users, Pulse Lab is in the process of user testing with gamers.

Read the full article on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

Indonesia drives SDG policy with a crowdsourced translation game was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

Contributed by Apolitical

This article was originally published on apolitical.co, a platform to help public servants tackle urgent challenges by connecting them to the most innovative policies and people worldwide.
—

Queensland public hospitals are saving $2.5 million (USD) a year by using a tool that predicts patient admissions — and their injury type — hours, weeks or even ten years in advance. The benefit to the state as a whole could be as high as $80 million per year from improved patient outcomes.

The tool uses algorithms to identify patterns in historic patient admission and discharge data, allowing hospitals to plan when to open and close beds, alter patient discharge plans and adjust staff vacations to create optimal patient flow rates. It is being used by around 50 hospitals in Queensland and has an accuracy rate of up to 95%.

Results & Impact

The financial return from implementing the Patient Admission Prediction Tool in Queensland is estimated to be as high as $80 million per year. This comprises $2.5 million in efficiency savings and $77.5 million from improved patient outcomes. If implemented across Australia it could save as much as $18.5 million per year from efficiency improvements.

Key Parties

• Commonwealth Scientific and Industrial Research Organisation
• Queensland Health

How

The Patient Admission Prediction Tool was first developed and tested by the Commonwealth Scientific and Industrial Research Organization in 2010 using patient admission and discharge data from two public hospitals. Regression analysis models were run to identify patterns in the records.

To test the system, scientists made predictions using historical data and compared the results with the actual number of patients admitted, calculating the percentage difference between the predicted and actual number of arrivals.

This system produced a model which could predict patient admissions and injury type with 90% to 95% accuracy on a day-to-day basis. The system was initially implemented in about 30 Queensland hospitals in 2011 and fully rolled out in 2014.

The tool has been integrated into a daily trigger warning system, whereby staff undertake a series of measures to free up bed space in response to specific levels of anticipated demand.

At the start of each day, the predicted number of arrivals is reviewed to see what demand for beds will be. If wait times are expected to rise above a certain level, a series of interventions is introduced to free up bed space, such as rescheduling elective procedures, adjusting discharge patterns or moving patients to different wards.

The tool is also used for scheduling operations, planning when to open or close beds and managing staffing levels on a longer term basis.

Hurdles

The Patient Admission Prediction Tool’s accuracy varies according to the time span of the prediction being run. While daily forecasts are 90% to 95% accurate, predictions for four-hour periods can be wrong by as much as 40% while hourly forecasts can have an error margin as high as 50%.

A second problem is the need to manage culture change. When implemented in Queensland, there was resistance from some hospital staff who were reluctant to adjust to new working patterns.

Read the full article on apolitical.co

—
If you liked this 👏 so other people will see it on Medium.

Connect with us: super.global | Twitter | LinkedIn | Instagram

Working on an impactful project that needs creative tech expertise? Find out how we’ll help you prepare for, find and collaborate with an outstanding agency partner.

Queensland hospitals can predict admissions ten years in advance was originally published in Super Global on Medium, where people are continuing the conversation by highlighting and responding to this story.

1. Grit

Around a hundred Londoners each year survive a stroke that would previously have killed them because the city created new specialist units in better locations. It’s a famous case of excellent governing making people’s lives better. But to manage it, the reformers had to de-fund opponents, overcome bitter resistance and close the best existing stroke unit in the city — the one used by parliament. They started planning in 2005, opened the first new unit in 2010 and got the medical results that conclusively vindicated them in 2014. The woman who oversaw it all, Ruth Carnall, was was a dame and the doctor who did her research, Ara Darzi, was made a Lord. They won, but the fight was long and hard.

2. Tech savvy

Big data may be the most boringly named revolution in history, but that doesn’t mean it isn’t going to become as ubiquitous and essential as electricity. From live-tracking the dampness of flower beds in Barcelona’s public parks to calculating the economic cost of crimes or analysing the knock-on effects of pedestrianising a road, data is everywhere. Add in blockchain and machine learning, and knowing the number for IT isn’t going to cut it any more.

3. Confidence

Arnoud Passenier, a Dutch public servant, who brought together companies and universities to try and clean up Rio de Janeiro’s polluted bay before last year’s Olympics, told Apolitical, ‘I wasn’t given this assignment. I created this with people in my network. Only a week before signing the agreement, I said, “Well, Minister and State Secretary, there are fifty-five parties involved, so please, would you like to sign the agreement?” Of course they did, because the plan fits their political agenda and saying no to such a large group is impossible.’

4. Collaboration

‘It’s amazing what you can accomplish if you do not care who gets the credit.’ So said President Truman, and it’s never been more true than today. The increasing importance and availability of networks, whether online or off, has multiplied the opportunities for drawing on other people’s skills. Britain’s Fast Stream for its future leaders now consciously tries to build up their personal networks by moving them around departments and into the private and non-profit sectors.

5. Internationalism

Every country in the world has cities, schools and hospitals, and almost every country has hundreds of thousands of public servants coming up with ways to lower air pollution, ease traffic, improve exam results and cut waiting times. Especially if you don’t have the budget for detailed research — and even if you do — it’s easier to copy someone else’s wheel than to reinvent it. Online platforms and Skype have made it possible.

6. The ability to listen

The world’s biggest source on data and experience of public services is the public itself, and effective civil servants are the ones who know how to tap it. In Britain, that’s one of the things underlying the big successes of its Nudge Unit, based on behavioural economics, which has done things like bring in hundreds of thousands of new organ donors by designing the ‘user experience’. In Sao Paulo, Brazil, 25,000 public servants a year are retrained by crowdsourcing the public’s expertise in things like data, management and technology.

7. Storytelling

To mobilise the people that your networks bring you into contact with, you need to be able to articulate a compelling vision of why it matters that they do what you want them to. Get your storytelling right and you can move mountains. Two public servants in the US, Gabby Dreyfus and Chad Gallinat, together set up something called the Global Lighting Challenge, which has now brought together 14 governments, the European Commission and major manufacturers like Philips and Osram to switch to energy-efficient LEDS.

8. Friends in business

Never has there been a better time to collaborate with the private and non-profit sectors, partly because government budgets are so constrained and partly because business is now ready for it. A third of social enterprises in the UK are less than three years old, compared with around 10% of other small businesses, and more of them are growing. In other words, social enterprise is where the action is, and 59% of them do business with government.

9. An appetite for risk

‘It’s very difficult to get fired in government, so it’s the place where it should be so easy to risk things,’ Virginia Hamilton, a pioneer of human-centred design in the US Labor Department, told Apolitical. ‘And you know, I don’t get paid enough to not have fun at work.’ Doing something new means sticking your neck out, and it might go wrong. If you don’t accept, calculate and mitigate that risk, you’ll either mess with ordinary people’s lives or never do anything to truly help them.

Finally, you also need a healthy dose of optimism. Working in government is difficult, complicated and sometimes frustrating, but it’s also the most powerful institution in the country, capable of doing things that no other organisation could imagine. It’s a great responsibility, but it’s great power, too.

• Do you want to get involved? Join our community of pioneers in public service or read our free weekly briefing.

African countries from Kenya to the Ivory Coast are learning the financial skills to stop some multinational oil and mining corporations taking advantage of their lack of know-how.

Dozens of developing countries sell access to oil, gold and other reserves without being able to analyse the financial basis on which these deals are made, and a new project to help them make the requisite financial models has started to reveal how much these deals are really costing them.

One of the first results, in the Ivory Coast, has shown that a poorly understood tax break for mining companies is costing the country hundreds of millions of dollars.

“The lack of capacity in the entire government is detrimentally shocking”

Poor management of abundant natural resources has created a phenomenon known as ‘the curse of oil’ or ‘the paradox of plenty’: countries rich in natural resources tend to have less growth, worse democracy and slower development than those without.

One official from a West African country, who did not want to be identified, told Apolitical, ‘The capacity in the entire government for financial modelling is detrimentally shocking. The main model supposedly used by the government — the Fari model developed by the IMF — is not kept up to date because nobody knows how to use it.

‘Every half year, officials from different ministries meet and update their entire macroeconomic forecasts on this basis, but no one really knows how the model works. This has led to tremendous mis-forecasting, which is one reason for the financial crisis. They forecast that they could spend a lot of money, which they have done, and now there are a lot of problems.

“Countries are making decisions about billions of dollars more or less blind”

‘Specialists come in, they give you a ready-made model, but then nobody knows how to use it because they don’t learn it from scratch. The Fari model might be good for the purposes of the IMF, but I suppose the person who developed it wasn’t thinking about teaching and educating people. Many people struggle with it. And people admit it themselves: it’s just too difficult. So this model sits around and, well, life goes on.’

In the new modelling project, officials from developing countries study financial modelling and analyze mines or oils fields in their countries. The project was set up by a company called Open Oil, whose founder said, ‘A large number of countries are making enormous decisions, about billions of dollars, more or less blind. We suspect that some are getting bad deals and a lot more simply don’t know what the terms are that they’ve signed.’

In the Ivory Coast, the project examined the five-year tax holiday given the Yaoure gold mine, a standard sweetener to attract foreign companies. Until now, no one in the Ivorian government has actually known how much that exemption was costing them. The new analysis calculated a cost of $110million dollars over five years. For the whole sector, it may be in the region of half a billion dollars. The average income in the Ivory Coast is around $3,100 a year.

“We were able to measure the shortfall for the state”

Bienvenu Esse, one of Ivorian officials who did the analysis, was ‘absolutely’ surprised by how much it was. ‘We quantified the financial impact and were able to measure the shortfall for the state. A review of all ongoing projects would certainly allow the adoption of strategies to improve the government’s revenues for future projects while developing a win-win policy with investors. The skills developed during this program will be deployed to better guide decisions to maximise gains for the state and local communities.’

Tax holidays and other concessions of this sort are given to make it worthwhile for foreign companies to invest — and the costs and risks may well be higher in Ivory Coast than other places.

‘It’s not for us to say it’s a good policy or a bad policy,’ said West. ‘But a second analysis they’ve done shows the profitability for the investor under both scenarios, with and without the tax holiday. And there we see that the rate of return, even without the tax holiday, is projected to be 23–24%. That is well above normally accepted industry minimums needed to justify an investment.’

“It is difficult but not impossible”

The project trains interested public servants to create the financial models — of projected investments, running costs and profits — on which these deals are based. Multinational companies make their own models but, according to West, at least two dozen countries that depend heavily on natural resources are unable to make similar models themselves. That means that when they offer terms to foreign companies they do not know the value of what they are proposing.

Public servants interested in financial modelling are helped by Open Oil to learn the necessary skills and make a model of a mine or oil field in their country, all of it with publicly available data. Most are doing it in their own time, and it is far from easy. West estimates that the whole process will take someone 300–500 hours, and that of the people who take part in his boot camp, one in ten make it through. Nevertheless, out of 300 people starting the course, there have been three or four with zero experience in the field who have now made professional-standard models. As West puts it, ‘We’ve proved that it is difficult but not impossible.’

A pivotal part of the program is international co-operation. Because financial models are so complex, they need to be reviewed by someone else, something made much more arduous by all the different ways in which these models can be constructed. Open Oil has developed a standard framework to make this peer review possible, and the participants look at one another’s models to help them solve the inevitable problems.

“You ask the group: who can help me?”

‘It’s very very useful,’ Anastasia Wanjohi, of Kenya’s Commission on Revenue Allocation, told Apolitical. ‘Because you forget some things and when you’re working on something, you never see mistakes, but when you have a network of modellers, you ask the group: who can help me? And someone will say, it’s this-and-this. It’s very very helpful and it brings checks into your model. When someone else looks at it, they can critique it constructively.’

Wanjohi and her colleague John Mose are modelling the Kwale mine, the biggest in Kenya, which produces the ores ilmenite, rutile and zircon. There have been local protests alleging that the country is not gaining enough from the mine, hence the need for financial modelling to actually determine the project’s economics.

Wanjohi and Mose are not yet ready to publish their results, but she said, ‘When investors come into the country, Kenya needs to understand exactly the models they bring, to be able to negotiate. Because every company, every investor, before they come to a decision to invest, they have the financial models. If government has the capacity to actually scrutinise them, it will be able to understand and question what exactly they are.’

“If you don’t understand, you can’t do anything”

The benefits of developing countries gaining this expertise could, if managed correctly, be cumulative, doing something to level up the balance of power between the owners of natural resources and those who work them. The Ivory Coast team want to model the entire sector, help set an internationally recognised standard and then require multinational companies to provide their own analyses in that format. There is also talk of using analyses of each country’s deals as benchmarks for the others.

This sort of ambitious expansion, which would ultimately allow countries like Ivory Coast to accurately predict how many bucks they will have in next year’s budget, relies on spreading this knowledge beyond a few pioneers. Otherwise, when they move on, the understanding will go with them. If that happens, these countries will be back where they started.

As Wanjohi put it, ‘If you don’t understand, you can’t question, you can’t do anything. That means we may get bad deals and be taken for a ride. But if we do understand, we’ll be able to get the best for Kenya and be economically empowered.’

(Picture via Flickr user JudyWeber33)

The MARIO project is run by a consortium of Irish, French, British, Italian and German developers, schools, nursing programs, robotics firms, social care programs and hospitals. MARIO received $4.3 million funding from Horizon2020 in 2015 to build Kompai-2 robots that improve dementia sufferers’ quality of life.

“Often, people with dementia tend to with withdraw. They stop participating in community life. That’s a problem, because isolation compounds dementia: the disease gets worse, and [their health] declines more quickly,” said Kathy Murphy, the communications officer for MARIO and a professor at the University of Ireland Galway.

In 2013, Alzheimer’s Europe estimated that there were 8.7 million people in the EU living with with dementia.

The robots are currently being tested in the UK, Ireland and Italy

“Looking at the funding situation and future increases in dementia, we realised that solutions weren’t going to come from expanding the number of nurses or carers. We wanted to look at more innovative solutions — and that’s why we developed MARIO. The idea came from working with people with dementia, and trying to look at what’s sustainable in the long term,” said Murphy.

MARIO robots can verbally interact with patients via voice-activated software, which supports dementia patients’ memory recall and cognitive skills. They have special applications designed to meet their users’ needs, such as a “My Hobbies” module with personalised music, news and games. The robots allow patients to live more independently by giving them the freedom to watch a movie or play a game without having to wait for a caregiver’s assistance.

“We even have representatives on our board with dementia”

The robots also have the ability to make the Comprehensive Geriatric Assessment, a test that measures an older person’s medical state and mental capacity. The assessment takes health care professionals about 30 minutes to complete, and having the robots do it will free up their time.

“MARIO is special because it has actually been driven by people with dementia — we even have representatives on our board with dementia. We’ve shaped MARIO around what people with dementia say they need, and what their carers say they need,” said Murphy.

Initially, the team behind MARIO found it difficult to reconcile different ideas how the robots should interact with patients. Robots professionals, software developers, nurses and academics all had unique perspectives and expectations for the project. By bringing dementia patients into the process and hearing about what types of care they needed, the team was able to narrow its goals.

Patients can use the robots to play films and music, and a social network is being developed

The other major challenge faced by the team was concern from dementia carers that MARIO robots would take their jobs. The project staff took great pains to work with them and explain their view that the initiative is not about replacing people — it’s about augmenting care, and supporting health professionals.

A pilot project began in August 2016, and will run until August 2017. Eleven robots are being trialled in different capacities in Ireland, the UK and Italy. The robots in Ireland are being tested with patients living in residential care, while the UK robots are installed in individual homes. The robots in Italy make the Comprehensive Geriatric Assessments.

According to Murphy, the pilot is focused on “acceptability” — whether people react favourably to living with robots. The results so far have been positive: “People with dementia like MARIO; they find it fun. We were surprised that more people did not find it odd having a robot around. But MARIO provided a real focus for grandparents and grandchildren — it gives them something very exciting to come in and visit.”

The MARIO robots are expected to be fully operational and ready to assist patients and caregivers by 2018, at which point organisers hope to bring the robots to market. They expect each MARIO model to cost from $5,400 to $7,550 to build. Project coordinators are currently working on building out new functionalities for the robot, including “My Community” and “My Social Network” features.

San Francisco, Los Angeles, Raleigh (North Carolina) and many others have started to post their restaurant hygiene ratings next to reviews on the hugely popular site, to prevent outbreaks of food poisoning and pressure dirty eateries into cleaning up.

‘We had a 15% pick up in users accessing the data in one week,’ Stephanie Cushing of the San Francisco Department of Public Health told Apolitical.

Previously, information about hygiene standards was only available on placards at the restaurants themselves or on little-visited government websites — while people looking at restaurant reviews wouldn’t know anything about the state of the kitchen.

Hygiene ratings are now displayed next to user reviews of the restaurant

The collaboration also cuts the other way — food safety inspectors are using the reviews, and a clever algorithm, to target which places they should be looking at.

Words and phrases like ‘dirty’ or ‘made me sick’ can be flagged to inspectors, so they visit those restaurants instead of making their spot checks at random. Merging these notifications with the history of hygiene violations and feeding all that information into an algorithm means it can predict the likelihood of problems at specific restaurants.

A study for the Harvard Business Review found that more than 80% of restaurants could be correctly classified into the top or bottom half of hygiene scores using only Yelp text and ratings.

This idea has been taken much further in Chicago, which has used all kinds of information — such as inspection scores, the age of the business and sanitation complaints — to predict which restaurants ought to be looked at.

The need for a more targeted system is pressing: Chicago has 42 employees trying to keep tabs on around 15,000 eateries.

Some 3,000 Americans die of food poisoning each year

The predictive analytics were bankrolled by a $1million grant from Bloomberg Philanthropies, set up by former New York mayor Michael Bloomberg. The city’s Department of Innovation and Technology developed the system with analysts from AllState Insurance and the Civic Consulting Alliance.

This analysis is not, however, available on any review sites, and nor are Chicago’s hygiene ratings. Although several cities have replicated the Yelp model, it has happened piecemeal. Moreover, several places have so far been unable to act on it because they don’t collect their information in the right format; some just have collections of PDFs.

The San Francisco variant, which has been running for seven years, now covers 5,000 restaurants and 2,000 other places where food is prepared. The information is now available is 17 municipalities, including Boulder, Colorado; Anchorage, Alaska; and the entire state of Florida.

A randomised controlled trial by Harvard Business Review found that restaurants whose low hygiene ratings are posted on Yelp tend to respond by raising standards — and perform better in subsequent inspections.

The website also specifies what hygiene violations have been recorded

Some 3,000 Americans die of food poisoning each year, according to the Centers for Disease Control and Prevention, and there are around 128,000 hospitalisations.

In several cities and other countries, local entrepreneurs have launched apps that draw on government data to provide simply the hygiene rating of any restaurant. The problem with these is that they themselves are little used. The most popular variant in the UK, for example, has been downloaded some 10,000 times from Google Play, while Yelp has been downloaded 50million times.

Overall, Yelp say they have around 25million users per month via the app; 72million via mobile and 77million via desktops (though it’s impossible to say how those groups overlap).

The collaboration with San Francisco’s health department has allowed Yelp to expand what it offers its readers, and CEO Jeremy Stoppelman said in a blog post, ‘Public/private partnerships like this don’t necessarily provide a direct contribution to Yelp’s bottom line, but evidence suggests [this] will have a positive impact on society.’ The company believes this model could also be expanded to other areas, such as healthcare.

Stefanie Cushing said, ‘We were originally approached by another company who tried to sell us a technology which would help me deploy inspectors using an app. But the Yelp proposition was different. We see it is valuable for our customers, who are the public.’