1
00:00:00,000 --> 00:00:06,200
Well now we're going to take a look at the FedEx critical tracking system, the demo that

2
00:00:06,200 --> 00:00:08,000
Adam and I were talking about earlier.

3
00:00:08,000 --> 00:00:14,560
So this is basically a nationwide view of all of our shipments and vehicles and as we

4
00:00:14,560 --> 00:00:19,320
drill down into that you'll notice that those bubbles expand into more clusters and what

5
00:00:19,320 --> 00:00:23,680
we can see here is in the Atlanta market we have 15 extra vehicles, that's what the blue

6
00:00:23,680 --> 00:00:29,440
means is we've got too many vehicles and 13 too few vehicles in Chattanooga so our planners

7
00:00:29,440 --> 00:00:32,200
may look at this and basically reposition vehicles accordingly.

8
00:00:32,200 --> 00:00:36,360
If we want to look at the activity that's currently going on on this, this is powered by Lifecycle

9
00:00:36,360 --> 00:00:40,400
Data Services in the background and we've got a continuous feed of what's going on with

10
00:00:40,400 --> 00:00:45,700
shipments and vehicles and as they move we get updates on this screen appropriately.

11
00:00:45,700 --> 00:00:49,680
Now when we drill into the Dallas market we can see that we've got a few extra vehicles

12
00:00:49,680 --> 00:00:50,680
here.

13
00:00:50,680 --> 00:00:54,580
We're going to drill down in one specific vehicle here and this is actually going to Atlanta

14
00:00:54,580 --> 00:00:58,880
and this is a pharmaceutical shipment where we need to maintain the temperature.

15
00:00:58,880 --> 00:01:04,800
We're going to shift over here to the tracking view and this is going to look at this specific

16
00:01:04,800 --> 00:01:08,840
shipment for its course and we're going to fast forward it for the purpose of this demo

17
00:01:08,840 --> 00:01:09,840
here.

18
00:01:09,840 --> 00:01:12,760
So this is actually replaying something that happened in the past.

19
00:01:12,760 --> 00:01:16,720
You'll notice that the trucks up here at the top we've got this funny looking red shape

20
00:01:16,720 --> 00:01:20,800
that basically represents a fence as we call it at FedEx.

21
00:01:20,800 --> 00:01:26,720
This is a fence that tells you that we're exactly 15 minutes away from this pickup location and

22
00:01:26,720 --> 00:01:28,760
we can associate many different events with fences.

23
00:01:28,760 --> 00:01:33,760
In this case the customer wanted to be notified when we're 15 minutes away from their dock door so

24
00:01:33,760 --> 00:01:35,760
they can get prepared for us.

25
00:01:35,760 --> 00:01:41,680
And the reason why the shape is funny is because if you traveled every possible road into this location

26
00:01:41,680 --> 00:01:44,720
this is exactly where you're 15 or 5 minutes away.

27
00:01:44,720 --> 00:01:50,720
In this case this is a one minute fence and we're going to associate an automatic arrival at the pickup location with that.

28
00:01:50,720 --> 00:01:54,680
So our drivers don't need to tell us that, you know, I basically arrived at the location.

29
00:01:54,680 --> 00:01:58,560
We can infer that by the position of the vehicle in our reasoning system that we have in the

30
00:01:58,560 --> 00:02:00,120
background.

31
00:02:00,120 --> 00:02:05,480
Now for this particular shipment we're maintaining a temperature of 2 to 8 degrees Celsius for

32
00:02:05,480 --> 00:02:10,040
the entire shipment and these are the remote probes that we have deployed in the trailer.

33
00:02:10,040 --> 00:02:15,000
So we're getting real time readings again through live cycle data services and this is the historical

34
00:02:15,000 --> 00:02:18,920
breadcrumb of the sensor data as it's happened over time.

35
00:02:18,920 --> 00:02:20,780
Now we've got something bad that happened.

36
00:02:20,780 --> 00:02:26,060
The temperature control unit, the TCU, has a status of off and it's red for some reason.

37
00:02:26,060 --> 00:02:26,860
So that's bad.

38
00:02:26,860 --> 00:02:31,100
We're going to continue monitoring this and make sure nothing else wrong happens.

39
00:02:31,100 --> 00:02:35,900
We've got a reasoning system in the background here that's basically tracking the temperature

40
00:02:35,900 --> 00:02:40,220
and the sensor reads over time and we're detecting that there's an upward trend.

41
00:02:40,220 --> 00:02:44,060
So this is another indication that somebody in our control center would basically view these

42
00:02:44,060 --> 00:02:48,460
abnormal events and monitor those and take proactive measures.

43
00:02:48,460 --> 00:02:52,140
So we'll notice that it's going up and it's getting close to the 8 degree range that we're

44
00:02:52,140 --> 00:02:53,740
supposed to maintain.

45
00:02:53,740 --> 00:02:56,060
So we're going to start to take corrective action here.

46
00:02:56,060 --> 00:02:59,660
We're going to contact the driver and figure out what's going on with this.

47
00:02:59,660 --> 00:03:04,860
Now our reasoning engine just told us that we're a degree and a half increase within just

48
00:03:04,860 --> 00:03:09,420
the last 30 minutes which is very abnormal and our temperature control unit is off.

49
00:03:09,420 --> 00:03:11,820
So at this point the driver is sending us a message.

50
00:03:11,820 --> 00:03:13,660
He says that I've blown a fuse in the unit.

51
00:03:13,660 --> 00:03:18,140
I need a spare because my spare failed and I need immediate repair.

52
00:03:18,140 --> 00:03:23,980
So our control center agent would find the nearest repair facilities that might have this fuse.

53
00:03:23,980 --> 00:03:25,820
So they'll bring up a reroute.

54
00:03:25,820 --> 00:03:29,340
So this is a dynamic thing that happens whenever something bad happens.

55
00:03:29,340 --> 00:03:29,740
Thermo King.

56
00:03:30,940 --> 00:03:34,060
And basically we'll find all of the repair facilities.

57
00:03:34,460 --> 00:03:37,100
In this case it's a Thermo King location in Birmingham.

58
00:03:37,660 --> 00:03:40,540
And we'll notice that it's going to add 22 minutes to our route.

59
00:03:40,540 --> 00:03:42,140
So let's select that route option.

60
00:03:42,620 --> 00:03:47,340
Now you'll notice that the gray line was the normal route that we originally were planned on.

61
00:03:47,340 --> 00:03:49,340
And the blue is the new route that we're going to go on.

62
00:03:50,140 --> 00:03:53,020
So now our reasoning engine is picking this up and saying,

63
00:03:53,020 --> 00:03:56,460
"Hey, you just detoured off the route that you're supposed to be on.

64
00:03:56,460 --> 00:03:57,660
You're 22 minutes late."

65
00:03:58,460 --> 00:03:59,100
So that's okay.

66
00:03:59,100 --> 00:04:02,540
We're going to continue to track that because we need to fix this temperature problem that's going on.

67
00:04:03,180 --> 00:04:05,740
Now we've got a fence at the facility there.

68
00:04:05,740 --> 00:04:08,780
And you'll notice that we can blend in satellite imagery as well.

69
00:04:09,580 --> 00:04:11,180
So if that's important.

70
00:04:11,180 --> 00:04:15,580
And we'll notice that we're getting another piece of data from our reasoning engine.

71
00:04:15,580 --> 00:04:18,140
It's telling us that our vehicle speed is decreasing.

72
00:04:18,140 --> 00:04:21,340
Which is what we'd expect since they're stopped at the repair facility.

73
00:04:21,340 --> 00:04:25,100
But in this case we're not on the planned route and the vehicle's decreasing speed.

74
00:04:25,100 --> 00:04:27,740
So somebody from a security perspective might be interested in that.

75
00:04:27,740 --> 00:04:30,780
Because we may have some very high value goods on the back of the truck

76
00:04:30,780 --> 00:04:33,260
that we need to make sure aren't being in danger.

77
00:04:34,140 --> 00:04:36,540
So now our temperature control unit is showing normal again.

78
00:04:36,540 --> 00:04:40,700
You'll notice that our sensor reads are going back into normal operating condition.

79
00:04:41,660 --> 00:04:44,860
And now we're going to progress on to the delivery location.

80
00:04:45,660 --> 00:04:49,900
And we've got fences conveniently located at the delivery location as well.

81
00:04:49,900 --> 00:04:53,900
So we can notify a delivery recipient that we're getting near their location.

82
00:04:56,140 --> 00:04:59,180
So as we continue to progress here, we'll look at the delivery location.

83
00:04:59,180 --> 00:05:04,540
You'll notice the scheduled arrival times there is 2:36 and scheduled departures 3:36.

84
00:05:04,540 --> 00:05:06,460
So we're going to spend 60 minutes at the delivery.

85
00:05:07,020 --> 00:05:09,180
When we go into the ETA timeline here,

86
00:05:09,740 --> 00:05:14,220
we'll notice that the shipment was on schedule for the majority of the shipment.

87
00:05:14,220 --> 00:05:18,860
And then we had this 22-minute spike where we added the reroute to fix the repair.

88
00:05:19,580 --> 00:05:25,580
And basically, this is okay because the promise time to the customer actually was,

89
00:05:25,580 --> 00:05:29,260
we delivered 22 minutes ahead of schedule for the promise time.

90
00:05:29,260 --> 00:05:31,580
And the actual delivery time was met.

91
00:05:32,060 --> 00:05:33,180
So the customer's happy.

92
00:05:33,740 --> 00:05:37,820
We weren't at risk for a claim, which could be a very high dollar claim for us

93
00:05:37,820 --> 00:05:40,780
if we're shipping high value perishable goods on the back of the truck.

94
00:05:40,780 --> 00:05:42,940
And we still met the customer requirements.

95
00:05:42,940 --> 00:05:44,940
So everybody's happy in the end.

96
00:05:44,940 --> 00:05:49,500
Having technology and capabilities like lifecycle data services helps us to visualize this

97
00:05:49,500 --> 00:05:55,340
in a very low latency ways that we can proactively take measures to fix situations.

98
00:05:55,340 --> 00:05:56,540
Can I just ask one question?

99
00:05:56,540 --> 00:05:56,860
Sure.

100
00:05:56,860 --> 00:06:01,180
About like thresholds as opposed to, like that sort of made it look as if you were,

101
00:06:01,180 --> 00:06:04,300
the operative was all, the agent was already looking at what was happening.

102
00:06:04,300 --> 00:06:04,940
Correct.

103
00:06:04,940 --> 00:06:09,820
So are you capable of setting a threshold so that if the temperature changes,

104
00:06:09,820 --> 00:06:11,740
that's when this kicks in?

105
00:06:11,740 --> 00:06:12,220
Absolutely.

106
00:06:12,220 --> 00:06:12,540
Okay.

107
00:06:12,540 --> 00:06:14,700
Yeah, so we're tracking moving averages.

108
00:06:14,700 --> 00:06:18,860
So we're tracking that the temperature increased by a spike of X amount.

109
00:06:18,860 --> 00:06:22,300
And if that happens over the course of some window of interest,

110
00:06:22,300 --> 00:06:26,380
so we're looking at the last 30 minutes in this case, a degree and a half in the last 30 minutes,

111
00:06:26,380 --> 00:06:29,020
that would have been the case where if somebody wasn't looking at this,

112
00:06:29,020 --> 00:06:31,580
they would have been notified that something bad's happening with this.

113
00:06:31,580 --> 00:06:36,860
And there is a more aggregate view that shows multi-shipment, multi-vehicle for this as well.

114
00:06:36,860 --> 00:06:41,740
And they get notified of those events, and then they can drill down into this perspective to do some diagnostics.

115
00:06:41,740 --> 00:06:44,220
Yeah, so this reasoning engine is your secret souls.

116
00:06:44,220 --> 00:06:44,620
Yes.

117
00:06:44,620 --> 00:06:47,580
And it's based on, I mean, all the rules are in our secret sauce,

118
00:06:47,580 --> 00:06:50,780
but it's based on the open source rules project from JBoss.

119
00:06:50,780 --> 00:06:54,540
And the mapping technology is ESRI's Flex API.

120
00:06:54,540 --> 00:06:55,580
Right.

121
00:06:55,580 --> 00:06:59,900
So another important feature, I think you asked before about why we selected Adobe,

122
00:06:59,900 --> 00:07:03,340
we can plop in third-party component libraries.

123
00:07:03,340 --> 00:07:05,340
So we can plop in the ESRI library.

124
00:07:05,340 --> 00:07:07,820
FedEx didn't write this mapping engine behind us.

125
00:07:07,820 --> 00:07:09,100
We don't want to do that, right?

126
00:07:09,100 --> 00:07:11,180
There are parties that specialize in that.

127
00:07:11,180 --> 00:07:17,100
So we can integrate and extend the Adobe platform by putting these together and mesh them together.

128
00:07:17,100 --> 00:07:17,420
Right.

129
00:07:17,420 --> 00:07:20,060
You can plug a lot of different things in there as time goes on.

130
00:07:20,060 --> 00:07:20,540
Yes.

131
00:07:20,540 --> 00:07:21,020
That's right.

132
00:07:21,020 --> 00:07:21,420
Well, great.

133
00:07:21,420 --> 00:07:22,700
Well, thanks for showing that demo.

134
00:07:22,700 --> 00:07:23,340
That was fun.

135
00:07:23,340 --> 00:07:24,380
Thank you, Michael.