ABOUT THE SPEAKER
Tapiwa Chiwewe - AI researcher
Tapiwa Chiwewe, PhD, manages the advanced and applied artificial intelligence group at IBM Research – Africa, which uses artificial intelligence to develop solutions for some of Africa's grand challenges whilst making scientific advances.

Why you should listen

An engineer and researcher of many interests, Tapiwa Chiwewe has worked in academia and industry, developing commercial products and conducting scientific research in many areas that include mining, healthcare, defense, astronomy and the environment.

Chiwewe studied at the University of Pretoria in South Africa where he earned a PhD in Computer Engineering. His roles have ranged from junior lecturer and researcher, software engineer, followed by an extended stint with South Africa’s Council for Scientific and Industrial Research.

More profile about the speaker
Tapiwa Chiwewe | Speaker | TED.com
TED@IBM

Tapiwa Chiwewe: You don't have to be an expert to solve big problems

Filmed:
1,397,285 views

Driving in Johannesburg one day, Tapiwa Chiwewe noticed an enormous cloud of air pollution hanging over the city. He was curious and concerned but not an environmental expert -- so he did some research and discovered that nearly 14 percent of all deaths worldwide in 2012 were caused by household and ambient air pollution. With this knowledge and an urge to do something about it, Chiwewe and his colleagues developed a platform that uncovers trends in pollution and helps city planners make better decisions. "Sometimes just one fresh perspective, one new skill set, can make the conditions right for something remarkable to happen," Chiwewe says. "But you need to be bold enough to try."
- AI researcher
Tapiwa Chiwewe, PhD, manages the advanced and applied artificial intelligence group at IBM Research – Africa, which uses artificial intelligence to develop solutions for some of Africa's grand challenges whilst making scientific advances. Full bio

Double-click the English transcript below to play the video.

00:12
One winter morning, a couple of years ago,
0
849
3857
00:16
I was driving to work
in Johannesburg, South Africa,
1
4730
2975
00:19
and noticed a haze hanging over the city.
2
7729
2301
00:22
I make that drive on most days,
3
10895
2095
00:25
so it was unusual
that I hadn't noticed this before.
4
13014
2873
00:28
Johannesburg is known
for its distinctive skyline,
5
16559
3089
00:31
which I could barely see that morning.
6
19672
1817
00:34
It didn't take long for me to realize
that I was looking at an enormous cloud
7
22322
4521
00:38
of air pollution.
8
26867
1198
00:40
The contrast between
the scenic environment I knew
9
28994
3571
00:44
and this smog-covered skyline
10
32589
2285
00:46
stirred up something within me.
11
34898
1603
00:49
I was appalled by the possibility
of this city of bright and vivid sunsets
12
37092
5476
00:54
being overrun by a dull haze.
13
42592
2134
00:57
At that moment, I felt an urge
to do something about it,
14
45258
4004
01:01
but I didn't know what.
15
49286
1656
01:03
All I knew was
I couldn't just stand idly by.
16
51808
3175
01:07
The main challenge was,
17
55564
2097
01:09
I didn't know much
about environmental science
18
57685
4237
01:13
air-quality management
19
61946
1525
01:15
or atmospheric chemistry.
20
63495
1423
01:17
I am a computer engineer,
21
65827
1889
01:19
and I was pretty sure I couldn't code
my way out of this air pollution problem.
22
67740
4019
01:23
(Laughter)
23
71783
1061
01:24
Who was I to do anything about this issue?
24
72868
2532
01:27
I was but a citizen.
25
75888
2544
01:31
In the following years,
I learned a very important lesson,
26
79786
3881
01:35
a lesson we all need to take to heart
if we are to work towards a better future.
27
83691
4158
01:40
Even if you're not an expert
in a particular domain,
28
88738
3674
01:44
your outside expertise may hold the key
29
92436
2763
01:47
to solving big problems
within that domain.
30
95223
2595
01:50
Sometimes the unique perspective you have
31
98381
2692
01:53
can result in unconventional thinking
that can move the needle,
32
101097
4461
01:57
but you need to be bold enough to try.
33
105582
3400
02:01
That's the only way you'll ever know.
34
109752
2182
02:04
What I knew back then
35
112889
1546
02:06
was that if I was even going
to try to make a difference,
36
114459
3952
02:10
I had to get smart
about air pollution first,
37
118435
3024
02:13
and so I became a student again.
38
121483
3095
02:17
I did a bit of basic research
39
125522
2106
02:19
and soon learned that air pollution
40
127652
2000
02:21
is the world's biggest
environmental health risk.
41
129676
3134
02:25
Data from the World Health Organization
42
133816
2374
02:28
shows that almost 14 percent
of all deaths worldwide in 2012
43
136214
5142
02:33
were attributable to household
and ambient air pollution,
44
141380
3684
02:37
with most occurring
in low- and middle-income countries.
45
145088
3459
02:41
Ambient air pollution alone
causes more deaths each year
46
149449
3731
02:45
than malaria and HIV/AIDS.
47
153204
2316
02:48
In Africa, premature deaths
from unsafe sanitation
48
156355
3833
02:52
or childhood malnutrition
49
160212
1739
02:53
pale in comparison
to deaths due to air pollution,
50
161975
3540
02:57
and it comes at a huge economic cost:
51
165539
2706
03:00
over 400 billion US dollars as of 2013,
52
168269
4133
03:04
according to a study by the Organisation
for Economic Cooperation and Development.
53
172426
4992
03:09
Now, in my work,
54
177862
3112
03:12
I explore new frontiers
for artificial intelligence,
55
180998
4186
03:17
where the symbiotic relationship
between man and machine
56
185208
3437
03:20
can find a beneficial footing
and help us to make better decisions.
57
188669
3912
03:25
As I thought about
the air pollution problem,
58
193249
3064
03:28
it became clear that we needed
to find a way to make better decisions
59
196337
3946
03:32
about how we manage air pollution,
60
200307
2633
03:34
and given the scale of the problem,
61
202964
1983
03:36
it was necessary to do it
in a collaborative way.
62
204971
2515
03:40
So I decided I'd better get to know
some people working within the field.
63
208310
4190
03:45
I started to speak to officials
from the City of Johannesburg
64
213500
3422
03:48
and other surrounding cities,
65
216946
1627
03:50
and I engaged the local
scientific community,
66
218597
2959
03:53
and I also made a few cold calls.
67
221580
2190
03:57
The process of engagement I embarked upon
68
225016
2553
03:59
helped me to develop
a deeper understanding of the problem.
69
227593
3273
04:03
It also helped me to avoid the trap
70
231510
2346
04:05
people in my profession sometimes
fall into when trying to innovate,
71
233880
3424
04:09
where we are quick to apply a technology
72
237328
2327
04:11
before we've firmly grasped
the problem at hand.
73
239679
2794
04:15
I began to develop an idea
74
243534
2138
04:17
about what I could do
to improve the situation.
75
245696
2468
04:20
I started by simply asking myself
76
248742
2416
04:23
how I could bring together
in some meaningful way
77
251182
2698
04:25
my skills in software engineering
and artificial intelligence
78
253904
3868
04:29
and the expertise of the people
I'd reached out to.
79
257796
2642
04:33
I wanted to create an online
air-quality management platform
80
261075
3993
04:37
that would uncover trends in pollution
81
265092
2081
04:39
and project into the future
82
267197
1926
04:41
to determine what outcomes
can be expected.
83
269147
2399
04:44
I was determined to see my idea
translate into a practical solution,
84
272462
5388
04:49
but I faced uncertainty
85
277874
2828
04:52
and had no guarantee of success.
86
280726
1901
04:55
What I had was a very particular set
of engineering skills,
87
283576
5318
05:00
skills I'd acquired over my career
88
288918
2436
05:03
(Laughter)
89
291378
1245
05:04
that were new to people who had
been working on the air pollution problem
90
292647
3462
05:08
for so many years.
91
296133
1518
05:09
What I have come to realize is that
sometimes just one fresh perspective,
92
297675
4969
05:14
one new skill set,
93
302668
1293
05:15
can make the conditions right
for something remarkable to happen.
94
303985
3421
05:20
Our willpower and imagination
are a guiding light,
95
308066
3691
05:23
enabling us to chart new paths
and navigate through obstacles.
96
311781
3737
05:28
Armed with a firmer understanding
of the air pollution problem,
97
316241
3658
05:31
and having managed to source
over a decade's worth of data
98
319923
3362
05:35
on air pollutant levels
99
323309
1452
05:36
and the meteorological conditions
for in and around Johannesburg,
100
324785
4474
05:41
my colleagues from South Africa
and China and myself
101
329283
4112
05:45
created an air-quality
decision support system
102
333419
2737
05:48
that lives in the cloud.
103
336180
1350
05:50
This software system
analyzes historical and real-time data
104
338210
4071
05:54
to uncover the spatial-temporal
trends in pollution.
105
342305
2668
05:57
We then used new
machine learning technology
106
345695
2855
06:00
to predict future levels of pollution
107
348574
2639
06:03
for several different pollutants
days in advance.
108
351237
2897
06:06
This means that citizens
can make better decisions
109
354849
3458
06:10
about their daily movements
110
358331
1715
06:12
and about where to settle their families.
111
360070
2015
06:14
We can predict adverse
pollution events ahead of time,
112
362575
3484
06:18
identify heavy polluters,
113
366083
1801
06:19
and they can be ordered
by the relevant authorities
114
367908
2523
06:22
to scale back their operations.
115
370455
2008
06:25
Through assisted scenario planning,
116
373315
2515
06:27
city planners can also make
better decisions
117
375854
2698
06:30
about how to extend infrastructure,
118
378576
2005
06:32
such as human settlements
or industrial zones.
119
380605
2787
06:36
We completed a pilot of our technology
120
384160
2745
06:38
that was run over a period of 120 days,
121
386929
3328
06:42
covering all of South Africa.
122
390281
2175
06:45
Our results were confirmed
123
393376
1819
06:47
when we demonstrated a tight correlation
124
395219
2563
06:49
between the forecasting data
125
397806
2397
06:52
and the data we were getting
on the ground.
126
400227
2595
06:55
Through our leadership,
127
403631
1670
06:57
we have brought cutting-edge,
world-leading assets
128
405325
3849
07:01
that can perform air-quality forecasting
129
409198
2666
07:03
at an unprecedented
resolution and accuracy,
130
411888
4120
07:08
benefiting the city that I drove into
one winter morning not very long ago,
131
416032
6849
07:14
and thought to myself,
132
422905
1531
07:16
"Something is wrong here.
I wonder what can be done?"
133
424460
2849
07:20
So here is the point:
134
428563
2341
07:23
What if I'd not investigated
the problem of air pollution further?
135
431895
3857
07:28
What if I'd not shown some concern
for the state of the environment
136
436951
3923
07:32
and just hoped that someone,
somewhere, was taking care of the matter?
137
440898
3976
07:37
What I have learned is that,
138
445858
1445
07:39
when embarking on a challenging endeavor
139
447327
2144
07:41
that advances a cause
that we firmly believe in,
140
449495
2638
07:44
it is important to focus
on the possibility of success
141
452157
3485
07:47
and consider the consequence
of not acting.
142
455666
3046
07:51
We should not get distracted
by resistance and opposition,
143
459426
3905
07:55
but this should motivate us further.
144
463355
2397
07:58
So wherever you are in the world,
145
466405
4507
08:02
the next time you find
146
470936
1439
08:04
that there's some
natural curiosity you have
147
472399
2194
08:06
that is being piqued,
148
474617
1600
08:08
and it's about something you care about,
149
476241
2055
08:10
and you have some crazy or bold ideas,
150
478320
2785
08:13
and perhaps it's outside
the realm of your expertise,
151
481129
3664
08:16
ask yourself this:
152
484817
1238
08:19
Why not?
153
487278
1237
08:22
Why not just go ahead
and tackle the problem
154
490349
2517
08:24
as best as you can, in your own way?
155
492890
2801
08:27
You may be pleasantly surprised.
156
495715
2254
08:31
Thank you.
157
499064
1151
08:32
(Applause)
158
500239
3649

▲Back to top

ABOUT THE SPEAKER
Tapiwa Chiwewe - AI researcher
Tapiwa Chiwewe, PhD, manages the advanced and applied artificial intelligence group at IBM Research – Africa, which uses artificial intelligence to develop solutions for some of Africa's grand challenges whilst making scientific advances.

Why you should listen

An engineer and researcher of many interests, Tapiwa Chiwewe has worked in academia and industry, developing commercial products and conducting scientific research in many areas that include mining, healthcare, defense, astronomy and the environment.

Chiwewe studied at the University of Pretoria in South Africa where he earned a PhD in Computer Engineering. His roles have ranged from junior lecturer and researcher, software engineer, followed by an extended stint with South Africa’s Council for Scientific and Industrial Research.

More profile about the speaker
Tapiwa Chiwewe | Speaker | TED.com