ABOUT THE SPEAKER
Mallory Freeman - Data activist
UPS's advanced analytics manager Mallory Freeman researches how to do the most good with data.

Why you should listen

Dr. Mallory Freeman is the Lead Data Scientist in the UPS Advanced Technology Group, working on research and development projects for UPS’s smart logistics network. She serves on the advisory board of Neighborhood Nexus, supporting data-driven insights for the greater Atlanta region.

Freeman earned her Ph.D. in industrial engineering from the Georgia Institute of Technology in 2014. Her thesis explored how to measure and improve humanitarian operations in practical ways -- with a special focus on the use of algorithms. While she was in graduate school, she helped lead supply chain optimization projects for the UN World Food Programme. 

Freeman earned her Master's in operations research from MIT and her Bachelor's in industrial and systems engineering from Virginia Tech. In her spare time, she enjoys cooking, travelling and volunteering her data skills.

More profile about the speaker
Mallory Freeman | Speaker | TED.com
TED@UPS

Mallory Freeman: Your company's data could help end world hunger

Filmed:
1,090,373 views

Your company might have donated money to help solve humanitarian issues, but you could have something even more useful to offer: your data. Mallory Freeman shows us how private sector companies can help make real progress on big problems -- from the refugee crisis to world hunger -- by donating untapped data and decision scientists. What might your company be able to contribute?
- Data activist
UPS's advanced analytics manager Mallory Freeman researches how to do the most good with data. Full bio

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

00:12
June 2010.
0
880
1760
00:15
I landed for the first time
in Rome, Italy.
1
3760
2880
00:19
I wasn't there to sightsee.
2
7800
1896
00:21
I was there to solve world hunger.
3
9720
3120
00:25
(Laughter)
4
13160
2096
00:27
That's right.
5
15280
1216
00:28
I was a 25-year-old PhD student
6
16520
2096
00:30
armed with a prototype tool
developed back at my university,
7
18640
3096
00:33
and I was going to help
the World Food Programme fix hunger.
8
21760
3080
00:37
So I strode into the headquarters building
9
25840
2736
00:40
and my eyes scanned the row of UN flags,
10
28600
2816
00:43
and I smiled as I thought to myself,
11
31440
1960
00:46
"The engineer is here."
12
34840
1616
00:48
(Laughter)
13
36480
2216
00:50
Give me your data.
14
38720
1776
00:52
I'm going to optimize everything.
15
40520
2176
00:54
(Laughter)
16
42720
1736
00:56
Tell me the food that you've purchased,
17
44480
1896
00:58
tell me where it's going
and when it needs to be there,
18
46400
2616
01:01
and I'm going to tell you
the shortest, fastest, cheapest,
19
49040
2736
01:03
best set of routes to take for the food.
20
51800
1936
01:05
We're going to save money,
21
53760
1496
01:07
we're going to avoid
delays and disruptions,
22
55280
2096
01:09
and bottom line,
we're going to save lives.
23
57400
2736
01:12
You're welcome.
24
60160
1216
01:13
(Laughter)
25
61400
1696
01:15
I thought it was going to take 12 months,
26
63120
1976
01:17
OK, maybe even 13.
27
65120
1560
01:19
This is not quite how it panned out.
28
67800
2280
01:23
Just a couple of months into the project,
my French boss, he told me,
29
71600
3776
01:27
"You know, Mallory,
30
75400
1816
01:29
it's a good idea,
31
77240
1656
01:30
but the data you need
for your algorithms is not there.
32
78920
3336
01:34
It's the right idea but at the wrong time,
33
82280
2536
01:36
and the right idea at the wrong time
34
84840
2296
01:39
is the wrong idea."
35
87160
1376
01:40
(Laughter)
36
88560
1320
01:42
Project over.
37
90960
1280
01:45
I was crushed.
38
93120
1200
01:49
When I look back now
39
97000
1456
01:50
on that first summer in Rome
40
98480
1656
01:52
and I see how much has changed
over the past six years,
41
100160
2656
01:54
it is an absolute transformation.
42
102840
2240
01:57
It's a coming of age for bringing data
into the humanitarian world.
43
105640
3400
02:02
It's exciting. It's inspiring.
44
110160
2656
02:04
But we're not there yet.
45
112840
1200
02:07
And brace yourself, executives,
46
115320
2296
02:09
because I'm going to be putting companies
47
117640
1976
02:11
on the hot seat to step up
and play the role that I know they can.
48
119640
3120
02:17
My experiences back in Rome prove
49
125520
2816
02:20
using data you can save lives.
50
128360
2080
02:23
OK, not that first attempt,
51
131440
2456
02:25
but eventually we got there.
52
133920
2576
02:28
Let me paint the picture for you.
53
136520
1736
02:30
Imagine that you have to plan
breakfast, lunch and dinner
54
138280
2736
02:33
for 500,000 people,
55
141040
1616
02:34
and you only have
a certain budget to do it,
56
142680
2136
02:36
say 6.5 million dollars per month.
57
144840
2240
02:40
Well, what should you do?
What's the best way to handle it?
58
148920
2762
02:44
Should you buy rice, wheat, chickpea, oil?
59
152280
2760
02:47
How much?
60
155760
1216
02:49
It sounds simple. It's not.
61
157000
2136
02:51
You have 30 possible foods,
and you have to pick five of them.
62
159160
3216
02:54
That's already over 140,000
different combinations.
63
162400
3416
02:57
Then for each food that you pick,
64
165840
1696
02:59
you need to decide how much you'll buy,
65
167560
1976
03:01
where you're going to get it from,
66
169560
1696
03:03
where you're going to store it,
67
171280
1480
03:05
how long it's going to take to get there.
68
173760
1976
03:07
You need to look at all of the different
transportation routes as well.
69
175760
3336
03:11
And that's already
over 900 million options.
70
179120
2080
03:14
If you considered each option
for a single second,
71
182120
2376
03:16
that would take you
over 28 years to get through.
72
184520
2336
03:18
900 million options.
73
186880
1520
03:21
So we created a tool
that allowed decisionmakers
74
189160
2456
03:23
to weed through all 900 million options
75
191640
2616
03:26
in just a matter of days.
76
194280
1360
03:28
It turned out to be incredibly successful.
77
196560
2240
03:31
In an operation in Iraq,
78
199400
1256
03:32
we saved 17 percent of the costs,
79
200680
2536
03:35
and this meant that you had the ability
to feed an additional 80,000 people.
80
203240
4136
03:39
It's all thanks to the use of data
and modeling complex systems.
81
207400
4400
03:44
But we didn't do it alone.
82
212800
1280
03:46
The unit that I worked with in Rome,
they were unique.
83
214840
2736
03:49
They believed in collaboration.
84
217600
1736
03:51
They brought in the academic world.
85
219360
1696
03:53
They brought in companies.
86
221080
1280
03:55
And if we really want to make big changes
in big problems like world hunger,
87
223200
3616
03:58
we need everybody to the table.
88
226840
2560
04:02
We need the data people
from humanitarian organizations
89
230040
2936
04:05
leading the way,
90
233000
1256
04:06
and orchestrating
just the right types of engagements
91
234280
2576
04:08
with academics, with governments.
92
236880
1696
04:10
And there's one group that's not being
leveraged in the way that it should be.
93
238600
3696
04:14
Did you guess it? Companies.
94
242320
2096
04:16
Companies have a major role to play
in fixing the big problems in our world.
95
244440
3600
04:20
I've been in the private sector
for two years now.
96
248880
2416
04:23
I've seen what companies can do,
and I've seen what companies aren't doing,
97
251320
3576
04:26
and I think there's three main ways
that we can fill that gap:
98
254920
3376
04:30
by donating data,
by donating decision scientists
99
258320
3096
04:33
and by donating technology
to gather new sources of data.
100
261440
3480
04:37
This is data philanthropy,
101
265920
1576
04:39
and it's the future of corporate
social responsibility.
102
267520
2840
04:43
Bonus, it also makes good business sense.
103
271160
2600
04:46
Companies today,
they collect mountains of data,
104
274920
3216
04:50
so the first thing they can do
is start donating that data.
105
278160
2762
04:52
Some companies are already doing it.
106
280946
2190
04:55
Take, for example,
a major telecom company.
107
283160
2416
04:57
They opened up their data
in Senegal and the Ivory Coast
108
285600
2776
05:00
and researchers discovered
109
288400
1976
05:02
that if you look at the patterns
in the pings to the cell phone towers,
110
290400
3334
05:05
you can see where people are traveling.
111
293758
1938
05:07
And that can tell you things like
112
295720
2176
05:09
where malaria might spread,
and you can make predictions with it.
113
297920
3096
05:13
Or take for example
an innovative satellite company.
114
301040
2896
05:15
They opened up their data and donated it,
115
303960
2016
05:18
and with that data you could track
116
306000
1656
05:19
how droughts are impacting
food production.
117
307680
2040
05:22
With that you can actually trigger
aid funding before a crisis can happen.
118
310920
3680
05:27
This is a great start.
119
315560
1280
05:29
There's important insights
just locked away in company data.
120
317840
2880
05:34
And yes, you need to be very careful.
121
322480
1816
05:36
You need to respect privacy concerns,
for example by anonymizing the data.
122
324320
3576
05:39
But even if the floodgates opened up,
123
327920
2776
05:42
and even if all companies
donated their data
124
330720
2536
05:45
to academics, to NGOs,
to humanitarian organizations,
125
333280
3256
05:48
it wouldn't be enough
to harness that full impact of data
126
336560
2976
05:51
for humanitarian goals.
127
339560
1520
05:54
Why?
128
342320
1456
05:55
To unlock insights in data,
you need decision scientists.
129
343800
3240
05:59
Decision scientists are people like me.
130
347760
2576
06:02
They take the data, they clean it up,
131
350360
1816
06:04
transform it and put it
into a useful algorithm
132
352200
2256
06:06
that's the best choice
to address the business need at hand.
133
354480
2840
06:09
In the world of humanitarian aid,
there are very few decision scientists.
134
357800
3696
06:13
Most of them work for companies.
135
361520
1640
06:16
So that's the second thing
that companies need to do.
136
364480
2496
06:19
In addition to donating their data,
137
367000
1696
06:20
they need to donate
their decision scientists.
138
368720
2160
06:23
Now, companies will say, "Ah! Don't take
our decision scientists from us.
139
371520
5736
06:29
We need every spare second of their time."
140
377280
2040
06:32
But there's a way.
141
380360
1200
06:35
If a company was going to donate
a block of a decision scientist's time,
142
383200
3416
06:38
it would actually make more sense
to spread out that block of time
143
386640
3136
06:41
over a long period,
say for example five years.
144
389800
2200
06:44
This might only amount
to a couple of hours per month,
145
392600
3056
06:47
which a company would hardly miss,
146
395680
2056
06:49
but what it enables is really important:
long-term partnerships.
147
397760
3480
06:54
Long-term partnerships
allow you to build relationships,
148
402920
2816
06:57
to get to know the data,
to really understand it
149
405760
2656
07:00
and to start to understand
the needs and challenges
150
408440
2416
07:02
that the humanitarian
organization is facing.
151
410880
2160
07:06
In Rome, at the World Food Programme,
this took us five years to do,
152
414345
3191
07:09
five years.
153
417560
1456
07:11
That first three years, OK,
that was just what we couldn't solve for.
154
419040
3336
07:14
Then there was two years after that
of refining and implementing the tool,
155
422400
3496
07:17
like in the operations in Iraq
and other countries.
156
425920
2800
07:21
I don't think that's
an unrealistic timeline
157
429520
2096
07:23
when it comes to using data
to make operational changes.
158
431640
2736
07:26
It's an investment. It requires patience.
159
434400
2400
07:29
But the types of results
that can be produced are undeniable.
160
437760
3496
07:33
In our case, it was the ability
to feed tens of thousands more people.
161
441280
3560
07:39
So we have donating data,
we have donating decision scientists,
162
447440
4336
07:43
and there's actually a third way
that companies can help:
163
451800
2696
07:46
donating technology
to capture new sources of data.
164
454520
2976
07:49
You see, there's a lot of things
we just don't have data on.
165
457520
2840
07:52
Right now, Syrian refugees
are flooding into Greece,
166
460960
2720
07:57
and the UN refugee agency,
they have their hands full.
167
465120
2560
08:01
The current system for tracking people
is paper and pencil,
168
469000
3056
08:04
and what that means is
169
472080
1256
08:05
that when a mother and her five children
walk into the camp,
170
473360
2856
08:08
headquarters is essentially
blind to this moment.
171
476240
2656
08:10
That's all going to change
in the next few weeks,
172
478920
2336
08:13
thanks to private sector collaboration.
173
481280
1880
08:15
There's going to be a new system based
on donated package tracking technology
174
483840
3656
08:19
from the logistics company
that I work for.
175
487520
2040
08:22
With this new system,
there will be a data trail,
176
490120
2336
08:24
so you know exactly the moment
177
492480
1456
08:25
when that mother and her children
walk into the camp.
178
493960
2496
08:28
And even more, you know
if she's going to have supplies
179
496480
2616
08:31
this month and the next.
180
499120
1256
08:32
Information visibility drives efficiency.
181
500400
3016
08:35
For companies, using technology
to gather important data,
182
503440
3256
08:38
it's like bread and butter.
183
506720
1456
08:40
They've been doing it for years,
184
508200
1576
08:41
and it's led to major
operational efficiency improvements.
185
509800
3256
08:45
Just try to imagine
your favorite beverage company
186
513080
2360
08:48
trying to plan their inventory
187
516280
1576
08:49
and not knowing how many bottles
were on the shelves.
188
517880
2496
08:52
It's absurd.
189
520400
1216
08:53
Data drives better decisions.
190
521640
1560
08:57
Now, if you're representing a company,
191
525800
2536
09:00
and you're pragmatic
and not just idealistic,
192
528360
3136
09:03
you might be saying to yourself,
"OK, this is all great, Mallory,
193
531520
3056
09:06
but why should I want to be involved?"
194
534600
1840
09:09
Well for one thing, beyond the good PR,
195
537000
2816
09:11
humanitarian aid
is a 24-billion-dollar sector,
196
539840
2776
09:14
and there's over five billion people,
maybe your next customers,
197
542640
3056
09:17
that live in the developing world.
198
545720
1816
09:19
Further, companies that are engaging
in data philanthropy,
199
547560
3096
09:22
they're finding new insights
locked away in their data.
200
550680
2976
09:25
Take, for example, a credit card company
201
553680
2256
09:27
that's opened up a center
202
555960
1336
09:29
that functions as a hub for academics,
for NGOs and governments,
203
557320
3376
09:32
all working together.
204
560720
1240
09:35
They're looking at information
in credit card swipes
205
563040
2736
09:37
and using that to find insights
about how households in India
206
565800
2976
09:40
live, work, earn and spend.
207
568800
1720
09:43
For the humanitarian world,
this provides information
208
571680
2576
09:46
about how you might
bring people out of poverty.
209
574280
2656
09:48
But for companies, it's providing
insights about your customers
210
576960
3016
09:52
and potential customers in India.
211
580000
2040
09:54
It's a win all around.
212
582760
1800
09:57
Now, for me, what I find
exciting about data philanthropy --
213
585960
3776
10:01
donating data, donating decision
scientists and donating technology --
214
589760
4336
10:06
it's what it means
for young professionals like me
215
594120
2376
10:08
who are choosing to work at companies.
216
596520
1840
10:10
Studies show that
the next generation of the workforce
217
598800
2656
10:13
care about having their work
make a bigger impact.
218
601480
2560
10:16
We want to make a difference,
219
604920
2456
10:19
and so through data philanthropy,
220
607400
2416
10:21
companies can actually help engage
and retain their decision scientists.
221
609840
3936
10:25
And that's a big deal for a profession
that's in high demand.
222
613800
2880
10:29
Data philanthropy
makes good business sense,
223
617840
3120
10:34
and it also can help
revolutionize the humanitarian world.
224
622200
3280
10:39
If we coordinated
the planning and logistics
225
627600
2096
10:41
across all of the major facets
of a humanitarian operation,
226
629720
3376
10:45
we could feed, clothe and shelter
hundreds of thousands more people,
227
633120
3600
10:49
and companies need to step up
and play the role that I know they can
228
637440
4256
10:53
in bringing about this revolution.
229
641720
1880
10:56
You've probably heard of the saying
"food for thought."
230
644720
2936
10:59
Well, this is literally thought for food.
231
647680
2240
11:03
It finally is the right idea
at the right time.
232
651560
4136
11:07
(Laughter)
233
655720
1216
11:08
Très magnifique.
234
656960
1576
11:10
Thank you.
235
658560
1216
11:11
(Applause)
236
659800
2851

▲Back to top

ABOUT THE SPEAKER
Mallory Freeman - Data activist
UPS's advanced analytics manager Mallory Freeman researches how to do the most good with data.

Why you should listen

Dr. Mallory Freeman is the Lead Data Scientist in the UPS Advanced Technology Group, working on research and development projects for UPS’s smart logistics network. She serves on the advisory board of Neighborhood Nexus, supporting data-driven insights for the greater Atlanta region.

Freeman earned her Ph.D. in industrial engineering from the Georgia Institute of Technology in 2014. Her thesis explored how to measure and improve humanitarian operations in practical ways -- with a special focus on the use of algorithms. While she was in graduate school, she helped lead supply chain optimization projects for the UN World Food Programme. 

Freeman earned her Master's in operations research from MIT and her Bachelor's in industrial and systems engineering from Virginia Tech. In her spare time, she enjoys cooking, travelling and volunteering her data skills.

More profile about the speaker
Mallory Freeman | Speaker | TED.com

Data provided by TED.

This site was created in May 2015 and the last update was on January 12, 2020. It will no longer be updated.

We are currently creating a new site called "eng.lish.video" and would be grateful if you could access it.

If you have any questions or suggestions, please feel free to write comments in your language on the contact form.

Privacy Policy

Developer's Blog

Buy Me A Coffee